JP2011047863A - Three-dimensional shape data processing apparatus, three-dimensional shape data processing system, and three-dimensional shape measurement system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To improve a scheme for determining whether alignment in the arrangement of shape data is proper for a target. <P>SOLUTION: The first shape data and the second-shape data expressing respective three-dimensional shapes of a first region and a second region having overlapping areas on the target are acquired; a plurality of third-shape data expressing a plurality of mutually different geometrical conditions are generated based on the second-shape data; and the respective third-shape data which have been generated are aligned in arrangement with the first-shape data, thereby generating a plurality of fourth-shape data. Whether the three-dimensional shape of the target can be reproduced with accuracy of a predetermined allowable range for a range containing the first region and the second region is determined, based on the correlation of the arrangement of the shape for the plurality of fourth-shape data. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a three-dimensional shape measurement system that generates shape data representing an overall three-dimensional shape of an object.

  In recent years, as one of the inspections of industrial products, the three-dimensional shape of parts of industrial products created based on design data such as CAD data is measured, and the measured three-dimensional shape and CAD data are measured using a computer. By comparison, CAT (Computer-Aided Testing) is performed to check whether or not the part is created with a shape error within a predetermined tolerance with respect to the design data.

  The three-dimensional shape measurement used in CAT is often a three-dimensional digitizer (also referred to as a “three-dimensional measuring machine”) that can measure the three-dimensional coordinates of many points on the surface of an object in a short time without contact. It is performed by three-dimensional measurement (three-dimensional measurement) using.

  The three-dimensional digitizer includes a passive type using a stereo image method and the like, an active type such as an optical radar method and an optical projection method, and an active passive type combining them. For example, in a three-dimensional digitizer using a stereo imaging method, a measurement object is photographed from a plurality of different directions using a plurality of cameras, and each of the measurement objects on the measurement object is obtained from the obtained images according to the principle of triangulation. Perform 3D measurement to calculate 3D coordinates of a point.

  In addition, a three-dimensional digitizer using a light projection method projects detection light from a projection device onto a measurement target, and receives reflected light from the measurement target with an imaging element of a camera. A slit light projection method (also called a light cutting method) uses slit light as detection light. In the slit light projection method, the slit light is deflected to optically scan the measurement object, and each point on the measurement object is determined by the principle of triangulation based on the degree of deformation of the slit light based on the surface shape of the measurement object. The three-dimensional measurement for calculating the three-dimensional coordinates is performed. A set of three-dimensional coordinates of each point is referred to as “three-dimensional shape data” or “shape data”.

  In this way, the three-dimensional digitizer can measure only the shape of the object facing the camera, the projection device, etc., and the shape of the object surface on the side not facing the three-dimensional digitizer is It cannot be measured. In addition, the range in which the 3D digitizer can perform the 3D measurement is a predetermined three-dimensional range (also referred to as “measurement range”) determined by the baseline length between the cameras, the angle of view of the camera, the depth of field, and the like. The measurement range is often smaller than the object.

  Therefore, in order to measure the overall (circumferential) shape of a part as an inspection object by CAT, by changing the position and orientation of the three-dimensional digitizer with respect to the part, a plurality of times from different directions can be obtained. Arrangement indicating the position and orientation of the 3D shape represented by each 3D shape data so that the obtained 3D shape data represents the overall 3D shape of the part. It is necessary to generate shape data for CAT that represents the overall shape of the component by adjusting each of the shapes (also referred to as “shape arrangement”).

  The adjustment of each shape arrangement for the plurality of three-dimensional shape data is an adjustment to match the plurality of three-dimensional shapes for the plurality of three-dimensional shape data with the original shape arrangement. Also referred to as “alignment” or “positioning”.

  There are several methods for alignment, but the first method is to perform three-dimensional measurement multiple times from different directions on the object so that the measurement ranges overlap on the object. Then, after obtaining a plurality of shape data representing the three-dimensional shape of each measurement range, for example, by using the ICP (Iterative Closest Points) method, a portion where the measurement ranges for each of the obtained plurality of shape data overlap There is a method of performing alignment for each piece of shape data based on the correspondence between the three-dimensional shapes. In patent document 1, the 1st method is employ | adopted.

  Further, as a second method of alignment, the same three-dimensional measurement as that of the first method is performed on the target object to which a plurality of markers are attached, and the target object is selected from the three-dimensional shapes of the target object. After obtaining two shape data respectively representing the three-dimensional shapes of the two measurement ranges having overlapping portions, the positions of the three or more markers existing in the overlapping portions of the two measurement ranges are representatively represented. “Marker position data” composed of a plurality of three-dimensional coordinates (also referred to as “three-dimensional positions”) is obtained for each shape data, and each shape data is determined based on the correspondence between the marker position data in each shape data. There is a method of performing the alignment of the above.

JP 2007-303839 A

  However, the ICP method employed in Patent Document 1 is an algorithm for obtaining coordinate conversion information that minimizes the distance between surfaces between three-dimensional shapes represented by a plurality of shape data by repeating iterative calculations. Therefore, in the case where a plurality of three-dimensional shapes having substantially the same shape are arranged by applying the ICP method, for example, a third order applying the ICP method such as planes or curved surfaces having a constant curvature. Depending on the features of the original shape, alignment may fail.

  FIG. 1 is a diagram illustrating an example of failure in alignment by the ICP method. In the example of FIG. 1, the two shapes are arranged so that the overlapping portions are in close contact with each other, but are arranged in the vertical direction (in the direction of the arrow Y <b> 1) so as to deviate from their intended positions.

  As described above, when substantially the same three-dimensional shape portion of the plurality of shape data obtained by measuring the part to be inspected is a shape unsuitable for alignment by the ICP method, the entire part to be inspected Shape data representing a typical three-dimensional shape cannot be generated accurately.

  That is, in the method of Patent Document 1, as a result of attempting to align the plurality of shape data, for example, the three-dimensional shapes of the three-dimensional shapes overlap with each other while deviating from the original arrangement. When there is almost no distance between the two surfaces, there is a problem that an alignment error cannot be detected depending on the mutual relationship between the shape data after the alignment.

  Also in the second alignment method, when all the three-dimensional positions of the markers existing in the overlapping portions of the plurality of measurement ranges are measured without error, all the three-dimensional positions corresponding to the markers are all Whether or not the alignment between shape data corresponding to multiple measurement ranges is correct can be determined based on whether or not they match, but for each shape data that is actually measured three-dimensionally, measurement errors during three-dimensional measurement Therefore, the three-dimensional position of each marker is measured by being shifted in various three-dimensional directions with respect to the actual three-dimensional position. Therefore, it is usually impossible for all three-dimensional positions corresponding to the markers to match each other. For this reason, there is a problem that it is difficult to determine whether or not the alignment is correct depending on the mutual relationship between the marker position data of each shape data after the alignment.

  Further, when the operator visually determines whether or not the alignment is correct, for example, the arrangement of the measurement range of the three-dimensional digitizer on the object at the time of three-dimensional measurement for acquiring a plurality of shape data to be determined By comparing this arrangement with the arrangement of the shape data after alignment displayed as a two-dimensional image on the display device by perspective projection or the like, it is allowable in the arrangement of the shape data. By verifying whether or not there is an impossible displacement, and determining whether or not the degree of overlap between the three-dimensional shapes is within an appropriate range based on the visual relationship between the displayed images Therefore, it is necessary to verify whether or not a microscopic misalignment has occurred.

  In order to efficiently perform these verifications, the perspective of the perspective projection is changed in various ways to change the verification part and the angle of the line of sight, and the display image is enlarged or reduced for full or partial verification. And it is necessary to change the colors and patterns to facilitate the distinction between the three-dimensional shapes, so it takes time for the determination process and determines whether or not alignment is possible. There is a problem that it is difficult to perform “processing” on a lot of shape data.

  In addition, since it is determined whether or not the alignment is possible without using an objective index, there are problems that the determination result varies depending on the skill level of the operator and the required time is long or short.

  In view of the above, an object of the present invention is to provide a technique for improving a technique for determining whether or not alignment is possible for three-dimensional shape data obtained by measuring a three-dimensional shape of an object.

  In order to solve the above-mentioned problem, the invention of claim 1 relates to a first area and a second area, which have portions overlapping each other on the object, in the three-dimensional shape of the object. An acquisition means for acquiring first shape data representing a three-dimensional shape and second shape data representing a three-dimensional shape of the second region, and a plurality of different geometries based on the second shape data Derivation data generating means for performing a derivation data generation process for generating a plurality of third shape data expressing a target state, and arranging each of the plurality of third shape data in the first shape data in a predetermined coordinate system By trying the alignment process, the shape arrangement adjusting means for generating a plurality of fourth shape data as a result of the process, and the arrangement of the first shape data and the second shape data in a single coordinate system Whether or not the three-dimensional shape of the object can be reproduced with accuracy within a predetermined allowable range for the range including the first area and the second area of the object. Determining means for determining based on the mutual relationship of the shape arrangement of the plurality of fourth shape data.

  Further, the invention of claim 2 is the first representation of the three-dimensional shape of the first region with respect to the first region and the second region having portions overlapping each other on the target among the three-dimensional shape of the target. An acquisition means for acquiring one shape data and second shape data representing a three-dimensional shape of the second region, and a plurality of different geometric states expressed based on the second shape data; Derived data generation means for performing derived data generation processing for generating third shape data, and shape arrangement adjustment for each of the plurality of third shape data to attempt alignment processing to the first shape data in a predetermined coordinate system And a range including the first region and the second region of the object when attempting to align the first shape data and the second shape data in a single coordinate system. , Whether the three-dimensional shape of the object can be reproduced with an accuracy within a certain allowable range, whether or not the plurality of second shape data when the plurality of third shape data is generated in the derived data generation processing The shape arrangement variation information and the plurality of shape arrangement variation information for the plurality of third shape data in the arrangement matching process are determined based on the interrelationships between the plurality of shape arrangement variation information respectively synthesized. Determining means for performing the above-described operation.

  The invention according to claim 3 is the three-dimensional shape data processing apparatus according to claim 1 or 2, wherein the plurality of geometric states are: (a) a three-dimensional shape in the predetermined coordinate system; At least one of the dislocation state of the shape arrangement corresponding to the general position and orientation and (b) the three-dimensional distortion state is a different state.

  The invention of claim 4 is the three-dimensional shape data processing apparatus according to any one of claims 1 to 3, wherein one of the plurality of third shape data is: It is the second shape data itself.

  Further, the invention of claim 5 is the three-dimensional shape data processing apparatus according to any one of claims 1 to 3, wherein all of the plurality of third shape data is the first shape data. It is characterized by expressing a geometric state different from the two-shape data.

  According to a sixth aspect of the present invention, there is provided a three-dimensional shape data processing device according to any one of the first to fifth aspects, and the first shape data used in the three-dimensional shape data processing device. And shape data generating means for generating the second shape data.

  The invention according to claim 7 is the three-dimensional shape data processing system according to claim 6, wherein the shape data generation means receives at least one of the first shape data and the second shape data as the It generates based on the known shape data which expressed the shape of the target object numerically by design.

  The invention of claim 8 is the three-dimensional shape data processing system according to claim 6, wherein the shape data generation means performs at least one of the first shape data and the second shape data as a third order. It produces based on the shape data about the said object measured using the measuring apparatus which measures original shape and acquires shape data, It is characterized by the above-mentioned.

  The invention according to claim 9 is the three-dimensional shape data processing system according to any one of claims 6 to 8, wherein the determination means denies the possibility of reproduction. When the result is obtained, (a) the shape data generating means generates fifth shape data representing a three-dimensional shape of the third region on the object different from any of the first region and the second region, The shape of the object is generated based on known shape data expressing the design numerically, and (b) the acquisition means acquires the fifth shape data instead of the second shape data. To do.

  The invention according to claim 10 generates shape data representing an overall three-dimensional shape of the object based on a plurality of individual shape data obtained by measuring the three-dimensional shape of the object from a plurality of directions. A three-dimensional shape measurement system, comprising: a measurement device that performs three-dimensional measurement to acquire the plurality of individual shape data; and an arrangement unit that relatively arranges the measurement device and the object. And a three-dimensional shape data processing system according to any one of claims 9 to 10, wherein the reproducibility is affirmed when the measuring device acquires the plurality of individual shape data. On the condition that the determination is obtained by the determination means, the range including the first region and the second region of the object is respectively included in the first shape data and the second shape data. Corresponding, Serial characterized by relatively disposed with said object and the measuring device on the basis of the measuring device and the positional relationship between the object.

  The invention according to claim 11 is the first region and the second region having three or more identification points in a portion overlapping each other on the target among the three-dimensional shape of the object provided with a plurality of identification points. And each of the three or more target identification points in the second region and the first position data representing the positions of three or more target identification points among the three or more identification points in the first region. Acquisition means for acquiring the second position data representing the position of the plurality, and based on the second position data, a plurality of states each representing a plurality of states having different distortion states of the distribution of the three or more identification points of interest Derived data generating means for generating derived data for generating third position data, and each of the plurality of third position data are subjected to an alignment process for the first position data in a predetermined coordinate system. The shape arrangement adjusting means for generating a plurality of fourth position data as a result of the processing, and the three-dimensional shape of the object with respect to the first region and the second region in a single coordinate system, respectively When trying to align the expressed shape data, the three-dimensional shape of the object with a precision within a predetermined allowable range with respect to a range including the first area and the second area of the object. And determining means for determining whether or not can be reproduced based on the mutual relationship of the shape arrangement of the plurality of fourth position data.

  The invention according to claim 12 is the first region and the second region having three or more identification points in a part of the three-dimensional shape of the object provided with a plurality of identification points on the target object. And each of the three or more target identification points in the second region and the first position data representing the positions of three or more target identification points among the three or more identification points in the first region. Acquisition means for acquiring the second position data representing the position of the plurality, and based on the second position data, a plurality of states each representing a plurality of states having different distortion states of the distribution of the three or more identification points of interest Derived data generating means for generating derived data for generating third position data, and each of the plurality of third position data are subjected to an alignment process for the first position data in a predetermined coordinate system. The object to be arranged when trying to align the shape data representing the three-dimensional shape of the object with respect to the first area and the second area in a single coordinate system. Of the first region and the second region, whether or not the three-dimensional shape of the object can be reproduced with an accuracy within a predetermined allowable range is determined by the plurality of second in the alignment process. Determination means for determining based on the mutual relationship of the variation information of the shape arrangement for each of the three position data.

  According to a thirteenth aspect of the present invention, there is provided the three-dimensional shape data processing apparatus according to the eleventh or twelfth aspect, and the first position data and the second position data used in the three-dimensional shape data processing apparatus. First shape data representing the three-dimensional shape of the object for the first region, and second shape data representing the three-dimensional shape of the object for the second region, used for each generation And a shape data generating means for generating.

  Further, the invention of claim 14 is based on a plurality of individual shape data obtained by measuring a three-dimensional shape of an object provided with a plurality of identification points from a plurality of directions. A three-dimensional shape measurement system for generating shape data representing a shape, wherein a measurement device that performs three-dimensional measurement to acquire the plurality of individual shape data, and the measurement device and the object are relatively arranged An arrangement unit and a three-dimensional shape data processing system according to claim 13, wherein when the measuring device acquires the plurality of individual shape data, the determination to affirm the possibility of reproduction is With respect to the range including the first region and the second region of the object on the condition that it is obtained by the determination unit, the range corresponds to the first shape data and the second shape data, respectively. Measurement Wherein the relative arrangement between the object and the measuring device on the basis of the positional relationship between the a location object.

  According to the first aspect of the present invention, the plurality of third shape data generated based on the second shape data representing the three-dimensional shape of the second region on the object has a portion overlapping the second region. The first shape data and the second shape data are based on the mutual relationship of the shape arrangement of the plurality of fourth shape data after being arranged with respect to the first shape data representing the three-dimensional shape of one region. Since it is determined whether or not the shape of the object can be reproduced with accuracy within a predetermined allowable range by arrangement, depending on the mutual relationship between the first shape data and the second shape data, the first shape data and the second shape Even when it is not possible to determine whether or not the data can reproduce the shape of the object with accuracy within a predetermined allowable range, it is possible to improve the probability of correctly determining whether or not alignment is possible.

  According to the eleventh aspect of the invention, it is generated based on the second position data representing the positions of three or more target identification points in the second region on the object provided with a plurality of identification points. A plurality of fourth position data generated by arranging the plurality of third position data to the first position data representing the positions of three or more target identification points in the first area having a portion overlapping the second area. Each shape data representing the three-dimensional shape of the object for the first region and the second region based on the mutual relationship of the shape arrangement for the position data is the three-dimensional shape of the object with a precision within a predetermined tolerance. Since it is determined whether or not the shape is reproducible, the three-dimensional shape of the object for the first region and the second region is determined depending on the mutual relationship between the shape positions of the first position data and the second position data. Improves the probability of correctly determining whether or not alignment is possible even if it is not possible to determine whether each represented shape data can reproduce the three-dimensional shape of the target with accuracy within a predetermined tolerance. Can be made.

It is a figure which shows the example of a failure of arrangement | positioning of the three-dimensional shape by ICP method. It is a block diagram which shows the structural example of the three-dimensional shape measurement system which concerns on embodiment, and a three-dimensional shape data processing system. It is a block diagram which illustrates the functional composition of the three-dimensional shape data processing device concerning an embodiment. It is a figure which shows the example of a setting of the measurement range with respect to a target object. It is a figure which shows the example of a setting of the measurement range with respect to a target object. It is a figure explaining acquisition of the individual shape data about a subject. It is a figure explaining acquisition of the individual shape data about a subject. It is a figure which shows a mode that two individual shape data were coordinate-transformed into 1 coordinate system. It is a figure which shows a mode that two separate shape data were arrange | positioned. It is a figure which shows the example of the different arrangement | positioning relationship between uneven | corrugated shape data. FIG. 11 is a diagram illustrating an example in which the alignment of the shape data illustrated in FIG. 10 is successful. It is a figure which shows the example which the alignment of shape data failed. It is a figure explaining acquisition of standard shape data and original reference shape data. It is a figure which illustrates standard shape data and original reference shape data. It is a figure which illustrates standard shape data and original reference shape data. It is a figure explaining a mode that the shape data for determination is produced | generated from CAD data. It is a figure which illustrates deviation distortion content reference shape data generated by giving distortion. It is a figure explaining a mode that the shift distortion content reference shape data generated by grant of arrangement gap is generated by three-dimensional measurement. It is a figure explaining a mode that the shift distortion content reference shape data generated by grant of arrangement gap is generated by three-dimensional measurement. It is a figure explaining the relationship between the original reference shape data and deviation distortion content reference shape data produced | generated by measurement. It is a figure explaining the fluctuation | variation of the apparatus arrangement | positioning of a measuring apparatus. It is a figure explaining a mode that the shift distortion content reference shape data to which arrangement gap was given is generated by operation. It is a figure explaining a mode that the shift distortion content reference shape data to which arrangement gap was given is generated by operation. It is a figure explaining a mode that standard shape data, original reference shape data, and shift distortion content reference shape data were changed into one coordinate system. It is a figure which illustrates the process in which the arrangement | positioning adjusted reference shape data are produced | generated. It is a figure which illustrates the arrangement | positioning adjusted reference shape data which arrangement | positioning succeeded. It is a figure which illustrates the arrangement | positioning adjusted reference shape data in which arrangement alignment failed. It is a figure which illustrates the process in which the arrangement | positioning adjusted reference shape data are produced | generated. It is a flowchart which shows the outline | summary of the example of a procedure which produces the shape data showing the whole three-dimensional shape of a target object. It is a flowchart which shows the outline | summary of the example of a procedure which produces | generates 2nd apparatus arrangement | positioning information. It is a flowchart which shows the outline | summary of the example of a procedure of arrangement | positioning permission determination processing. It is a block diagram which shows the structural example of the three-dimensional shape measurement system which concerns on a modification, and a three-dimensional shape data processing system. It is a block diagram which illustrates the functional structure of the three-dimensional shape data processing apparatus which concerns on a modification. It is a figure which shows the example of a setting of the measurement range with respect to the target object to which the marker was stuck. It is a figure which illustrates the individual shape data of the target object with which the marker was stuck. It is a figure which illustrates the individual shape data arranged based on the marker. It is a figure explaining the error of arrangement alignment resulting from a marker position shape. It is a figure explaining the error of arrangement alignment resulting from a marker position shape. It is a figure explaining the error of arrangement alignment resulting from a marker position shape. It is a figure explaining the error of arrangement alignment resulting from a marker position shape. It is a figure explaining the example of a production | generation of the reference | standard shape data and original reference shape data about the target object to which the marker was stuck. It is a figure explaining the example of a production | generation of the reference | standard shape data and original reference shape data about the target object to which the marker was stuck. It is a figure explaining the operation example of a marker position data acquisition part and a corresponding relationship acquisition part. It is a figure explaining the example of a production | generation of distortion containing reference marker position data. It is a figure explaining the production | generation process of arrangement | positioning adjusted marker position data which arrangement | positioning succeeded. It is a figure explaining the production | generation process of arrangement | positioning adjusted marker position data in which arrangement alignment failed. It is a flowchart which shows the outline | summary of the example of a procedure of arrangement | positioning permission determination processing.

<About Embodiment 1:>
[Three-dimensional shape measurement system 500A and three-dimensional shape data processing system 300A:]
FIG. 2 is a block diagram illustrating a configuration example of the three-dimensional shape measurement system 500A and the three-dimensional shape data processing system 300A according to the first embodiment.

  The three-dimensional shape processing apparatus 100A, the three-dimensional digitizer 2, the manipulator 3a, the manipulator control unit 3b, and the CAD system 8 shown in FIG. 2 form the main parts of the three-dimensional shape measurement system 500A and the three-dimensional shape data processing system 300A. In some cases, it functions as the three-dimensional shape measurement system 500A, and it functions as the three-dimensional shape data processing system 300A.

  In the three-dimensional shape data processing system 300A, the three-dimensional digitizer 2 arranged based on a plurality of first apparatus arrangement information indicating the position and orientation of the three-dimensional digitizer 2 with respect to the base portion of the manipulator 3a measures the object 1. And a plurality of shape data (referred to as “determination shape data”) each representing a three-dimensional shape substantially the same as a plurality of shape data to be obtained if measured from a plurality of directions so as to overlap. The shape data for the determination sets, for example, a plurality of sets of corresponding points dynamically or statically between the three-dimensional shapes to be aligned, such as an ICP method and a spin image method, which will be described later. It is possible to align whether or not it can be properly aligned by the “alignment method based on shape correspondence points” that performs alignment based on Determined by the determination process (described later), and generates a second device arrangement information first device arrangement information of a plurality each corresponding to each determination shape data it is determined to be disposed aligned.

  The three-dimensional shape measurement system 500A functions as the three-dimensional shape data processing system 300A in advance to generate a plurality of second device arrangement information, and then based on the generated plurality of second device arrangement information 1 and a three-dimensional digitizer 2 are relatively arranged, and a plurality of shape data (referred to as “individual shape data”) obtained by measuring the three-dimensional shape of the object 1 from a plurality of directions are shaped by an ICP method or the like. Shape data representing the overall three-dimensional shape of the object 1 is generated by applying an alignment method based on corresponding points.

  In the present embodiment, the three-dimensional shape measurement system 500A and the three-dimensional shape data processing system 300A are configured by the same components, but the three-dimensional shape measurement system and the three-dimensional shape data processing system are, for example, The three-dimensional digitizer and the manipulator having the same performance, and a separate component including the control device thereof may be configured. For example, for the manipulator and the control device of the three-dimensional shape data processing system 300A, You may employ | adopt a thing with higher arrangement determination precision than the thing of the measurement system 500A.

  When it is necessary to generate shape data used for shape inspection for many parts having the same shape, first, the three-dimensional shape data processing system 300A uses a plurality of parts for the first part. Second apparatus arrangement information is obtained, and then a plurality of second apparatuses obtained for other parts installed in substantially the same arrangement as the first part with respect to the base part of the manipulator 3a. If 3D measurement is performed by the 3D shape measurement system 500A using the arrangement information, individual shape data that can be properly arranged for other parts can be measured, and the time required for shape measurement for inspection of many parts can be reduced. can do. The operation of acquiring a plurality of pieces of second apparatus arrangement information based on the first part here is also referred to as “teaching” or the like.

  Next, main parts constituting the three-dimensional shape measurement system 500A and the three-dimensional shape data processing system 300A will be described.

○ 3D digitizer 2:
The three-dimensional digitizer 2 is a three-dimensional digitizer that measures the three-dimensional shape of the object 1 and acquires shape data representing the three-dimensional shape, and corresponds to the “measuring device” of the present invention. The three-dimensional digitizer 2 is connected to the three-dimensional shape data processing device 100A (FIG. 3), performs three-dimensional measurement of the object 1 in accordance with a control command from the three-dimensional shape data processing device 100A, and uses the measured shape data. Necessary for supplying to the three-dimensional shape data processing apparatus 100A and for use in extracting a marker or the like affixed on an object by image processing, as in an embodiment according to a modified example to be described later. Accordingly, the measured image data of the object is supplied to the three-dimensional shape processing apparatus 100A. In general, the three-dimensional digitizer 2 can change the measurement distance indicating the distance to the object at the time of measurement, and can be obtained by increasing the measurement range on the object as the measurement distance becomes longer. Conversely, when the resolution of the shape data is reduced and the measurement distance is shortened, the resolution of the obtained shape data is increased.

  When the three-dimensional digitizer 2 functions as each component of the three-dimensional shape data processing system 300A and the three-dimensional shape measurement system 500A, the three-dimensional digitizer 2 uses a plurality of pieces of shape data obtained by the three-dimensional measurement. Is supplied to the three-dimensional shape processing apparatus 100A as shape data for determination d1 (FIG. 3) or a plurality of individual shape data d2 (FIG. 3).

  Here, the shape data measured by the three-dimensional digitizer 2 is usually continuous based on the same measurement parameter in the same arrangement due to an error in the control of the measurement apparatus and noise included in the measured image data. Even if three-dimensional measurement is performed, slight variations occur for each measurement.

  In the 3D digitizer 2, calibration is performed based on a calibration object whose shape is known in order to accurately perform 3D measurement. This calibration residual, the attitude of the apparatus, and the temperature drift during use The error of three-dimensional measurement also occurs due to the above.

  In general, the 3D digitizer 2 defines “measurement accuracy” representing the accuracy of 3D measurement based on the errors and variations of the 3D measurement described above. It is necessary to pay attention to. In addition, the measurement accuracy is normally defined for each of several measurement conditions determined by, for example, a measurement distance.

  The measurement accuracy d18 of the three-dimensional digitizer 2 is supplied to the three-dimensional shape processing apparatus 100A from the three-dimensional digitizer 2 or input from the operation unit 4 (FIG. 3) or the like, so that the storage unit 14 (FIG. 3 ) And used as needed.

  Further, as the three-dimensional digitizer 2, in addition to the above-described type in which image data of the measured object is processed therein to generate shape data, the measured object image data is converted into the three-dimensional shape data processing apparatus 100A. The conversion from the image data to the shape data may be a type performed in the three-dimensional shape data processing apparatus 100A. In this case, the three-dimensional digitizer 2 and the three-dimensional shape processing apparatus 100A constitute the measuring apparatus of the present invention.

  In addition, as a measuring apparatus according to the invention that does not perform three-dimensional measurement of an object with a marker attached among the inventions of the present application, for example, a laser tracker that performs three-dimensional measurement by a method different from triangulation based on images Can be adopted.

○ Manipulator 3a and manipulator controller 3b:
The manipulator 3a includes an articulated arm portion and a pedestal portion on which the arm portion is installed. The manipulator 3a can rotate the arm part with respect to the pedestal part and bend and extend the arm part by the control from the manipulator control part 3b. Can be determined at an arbitrary position in a space within a predetermined range centered on the screen with a posture within the predetermined range. A three-dimensional digitizer 2 is attached to the tip of the arm portion.

  The manipulator control unit 3b is connected to the three-dimensional shape data processing device 100A via a connection cable. The three-dimensional shape data processing device 100A receives various control commands and the three-dimensional digitizer 2 for the pedestal portion of the manipulator 3a. Device arrangement information d4 (FIG. 3) indicating the position and orientation is supplied to the manipulator control unit 3b.

  The manipulator control unit 3b controls the operation of the manipulator 3a so that the three-dimensional digitizer 2 is arranged in a predetermined arrangement based on the apparatus arrangement information d4. The input of the device arrangement information d4 to the manipulator control unit 3b can be performed by the measurer inputting from an input unit (not shown) provided in the manipulator control unit 3b, and the operation of the manipulator 3a is also as follows. It can be controlled from this input.

  Normally, the device arrangement information d4 is based on the object arrangement information d20 indicating the position and orientation of the object with respect to the pedestal of the manipulator 3a stored in the storage unit 14 (FIG. 3). Thus, the three-dimensional digitizer 2 is set so as to be arranged in a desired relative arrangement. Therefore, the three-dimensional digitizer 2 is determined based on the device arrangement information d4 set based on the relative arrangement relationship between the object 1 and the three-dimensional digitizer 2.

  When the placement of the 3D digitizer 2 based on the device placement information d4 is finished, the manipulator control unit 3b sends the device placement information d4 to the 3D shape processing device in order to confirm the placement decision and to confirm the device placement information d4. Feedback to 100A. When the three-dimensional shape processing apparatus 100A recognizes that the arrangement of the three-dimensional digitizer 2 has been completed, the three-dimensional shape processing apparatus 100A causes the three-dimensional digitizer 2 to perform three-dimensional measurement. As for the device arrangement information d4, the three-dimensional shape of the three-dimensional digitizer 2 itself and the three-dimensional shape of the device arrangement detection structure attached to the three-dimensional digitizer 2 are measured using another shape measuring instrument. It is also possible to acquire the output by providing a sensor capable of detecting the device arrangement information d4 in the three-dimensional digitizer 2 and acquiring the output.

  Here, in the manipulator 3a, when determining the position and orientation of the three-dimensional digitizer 2 attached to the tip of the arm unit, an error in determining the arrangement according to the weight of the attached one or the like occurs. “Positioning accuracy” representing the certainty of placement is defined.

  The placement accuracy d19 of the manipulator 3a is supplied in advance to the three-dimensional shape processing apparatus 100A from the manipulator control unit 3b, or input from the operation unit 4 (FIG. 3) or the like, so that the storage unit 14 (FIG. 3) in advance. And used as needed.

  In addition to the configuration shown in FIG. 3, a configuration in which the object 1 is attached to the arm portion of the manipulator 3 a and the object 1 is moved with respect to the fixed three-dimensional digitizer 2 may be employed. .

  The manipulator 3a and the manipulator control unit 3b correspond to the “placement unit” of the present invention. As a placement means, in addition to the manipulator, a CMM (Coordinate-measuring machine) contact type probe that measures the shape of the object with a contact type probe is removed, and a mount part to which a three-dimensional digitizer 2 or the like can be attached is provided. Various arrangement determining mechanisms that can set the three-dimensional digitizer 2 and the like to a plurality of arrangements, such as those provided, may be employed.

  Further, as the arrangement means, various arrangement determining mechanisms such as a belt conveyor and a rotary stage that can set the object in a plurality of arrangements may be adopted, and a plurality of three-dimensional digitizers 2 and objects 1 may be used. Various arrangement determining mechanisms that can be set for the arrangement of the above may be employed together.

  Note that when the object 1 is moved, the arrangement information of the object 1 set based on the relative arrangement relationship between the object 1 and the three-dimensional digitizer 2 is the same as when the three-dimensional digitizer 2 is moved. Based on the above, the placement means determines the placement of the object 1.

  That is, the arrangement means of the present invention is an arrangement means for relatively arranging a measuring device such as the three-dimensional digitizer 2 and an object.

○ CAD system 8:
The CAD system 8 creates or accumulates CAD data that is information representing the design three-dimensional shape of various mechanical devices, products, practical products, and other articles. Further, the CAD data may include characteristic information such as the material and material of each article, the surface roughness of each part, and dimensional tolerance in addition to the information representing the design three-dimensional shape. The CAD system 8 is connected to the three-dimensional shape data processing apparatus 100A via a communication line or the like.

  When the three-dimensional shape data processing system 300A is functioning, the three-dimensional shape processing apparatus 100A has the CAD data corresponding to the object identification information such as a part number input from the operation unit 4 (FIG. 3). Requesting the supply, the CAD system 8 supplies the CAD data d5 (FIG. 3) according to the request to the three-dimensional shape processing apparatus 100A. Here, the CAD data d5 may be data in parts, or may be data in which one part is divided for each part that is a characteristic part of the part.

[Overview of Principle of Arrangement Acceptability Determination Processing According to Embodiment 1]
Next, prior to the description of the three-dimensional shape processing apparatus 100A, the outline of the principle of the alignment possibility determination process according to the first embodiment will be described together with the outline of the operation of the three-dimensional shape measurement system 500A.

○ Outline of operation of the three-dimensional shape measurement system 500A:
4 and 5 are diagrams for explaining setting of a measurement range when the three-dimensional shape measurement system 500A measures the three-dimensional shape of an object from a plurality of directions. In FIG. 4, the object 1 is three-dimensionally measured from a plurality of directions in which the three-dimensional digitizer 2 determined based on the device arrangement information d4a and d4b is different. Further, three-dimensional measurement may be performed by combining a plurality of three-dimensional digitizers such as the different three-dimensional digitizers 2a and 2b shown in FIG. Thus, depending on the limitation of the manipulator's movable range, some manipulators and measuring devices cannot measure all three-dimensional shapes of the target portion of the object, or measurement is necessary due to the movement time of the manipulator. When the time exceeds a predetermined time limit, a plurality of measuring devices may be used in combination with a manipulator. Further, the entire three-dimensional shape of the object may be measured by placing the object on a rotary stage or the like.

  Here, the measurement ranges n1 and n2 indicate the measurement ranges of the three-dimensional digitizer 2 corresponding to the device arrangement information d4a and the device arrangement information d4b, respectively, and represent the three-dimensional shapes corresponding to the measurement ranges n1 and n2, respectively. The data is supplied to the three-dimensional shape processing apparatus 100A as a plurality of individual shape data d2 (FIG. 3). The measurement ranges n1 and n2 have overlapping portions on the object 1, and the overlapping portions are provided with cylindrical and quadrangular columnar portions. The plurality of individual shape data d2 corresponding to the measurement ranges n1 and n2 are arranged based on the cylindrical and quadrangular column portions by the three-dimensional shape processing apparatus 100A using the ICP method or the like. Thus, when using the alignment method based on the shape corresponding points such as the ICP method, the measurement range is such that the uneven shape exists in the overlapping portion of the measurement range for obtaining the shape data to be aligned. Need to be set.

  In the case where there is no characteristic shape that forms irregularities such as these cylindrical and quadrangular columnar portions in the overlapping portion, and the overlapping portion is, for example, a flat surface, it is based on a shape corresponding point such as the ICP method. Even if the alignment method is applied, correct alignment is not performed. In such a case, for example, alignment based on a marker described later is employed.

  FIG. 5 shows an example of setting the measurement range for the quadrangular pyramid-shaped object 1a, and the measurement ranges n1 to n4 having overlapping portions are set clockwise with no gaps. The plurality of individual shape data d2 corresponding to each measurement range is aligned by the ICP method or the like mainly based on the cylindrical portion provided in the overlapping portion. When it is necessary to acquire shape data that is not missing, it is necessary to set the measurement range without any gaps in the target portion such as the inspection target of the object.

  6 and 7 are diagrams for explaining acquisition of individual shape data for an object. In FIG. 6, the three-dimensional shape of the object 1 formed by the curved surface with unevenness is sequentially arranged based on the device arrangement information d4a and the device arrangement information d4b along the arrow Y2, and the measurement range n1 in each arrangement. And a three-dimensional digitizer 2 having n2.

  Here, since part of the measurement ranges n1 and n2 overlaps on the object 1, the measured shape data has shape data representing substantially the same three-dimensional shape.

  FIG. 7 shows the individual shape data d2a and d2b determined based on the device arrangement information d4a and the device arrangement information d4b and measured by the three-dimensional digitizer 2 shown in FIG. 6 as the three-dimensional shapes they represent. ing.

  FIG. 8 is a diagram illustrating a state in which the individual shape data d2a and d2b illustrated in FIG. 7 are coordinate-converted into one coordinate system. 8 to 12, the amount of displacement between the three-dimensional shapes is exaggerated with respect to the size of the three-dimensional shape.

  This coordinate conversion can be performed based on the device arrangement information d4a and d4b. That is, the apparatus arrangement information d4a and d4b and the actual arrangement of the three-dimensional digitizer 2 decided based on the apparatus arrangement information d4a and d4b have errors due to the arrangement error of the manipulator 3a described above. Because of this error, as shown in FIG. 8, the individual shape data d2a and d2b usually do not overlap.

  FIG. 9 is a diagram showing a state in which the two individual shape data d2a and d2b whose arrangement shown in FIG. 8 is shifted are arranged, and this arrangement is a three-dimensional representation represented by the individual shape data d2a and d2b, respectively. Based on shape portions having substantially the same shape, an alignment method based on shape corresponding points such as an ICP method and a spin image method is used.

  Here, multiple sets of corresponding points are set between 3D shape data by converting each of multiple 3D shape data to be aligned into spin images and searching for multiple sets of corresponding points between each spin image. In this application, a technique for performing alignment between three-dimensional shape data is referred to as “spin image method”.

  The “spin image” is a predetermined feature that does not depend on the arrangement of the three-dimensional shape data for each point of the three-dimensional shape data, based on the three-dimensional positional relationship between each point and the surrounding points. It is an image obtained by specifying the amount.

  Since the spin image does not depend on the arrangement of the three-dimensional shape data, in the spin image method, the correspondence between the three-dimensional shape data is normally obtained statically and the arrangement between the three-dimensional shape data is performed.

○ Outline of Principle of Arrangement Acceptability Determination Processing in Embodiment 1:
Next, the ICP method will be described using an example in which the ICP method is applied to two shape data to perform alignment between the shape data. Here, the three-dimensional shapes represented by the two shape data respectively include substantially the same three-dimensional shapes, and are shifted from each other as in the individual shape data d2a and d2b shown in FIG. It shall be.

  The ICP method obtains a correspondence relationship between the reference shape data serving as a reference for alignment for each calculation and the reference shape data for alignment, and the surface of the three-dimensional shape represented by each shape data based on the correspondence relationship. The coordinate conversion information for shortening the distance between the surface and the surface is obtained.

  Coordinates that minimize the distance between surfaces between each three-dimensional shape by repeatedly calculating the coordinate conversion information and applying the coordinate conversion based on the obtained coordinate conversion information to the reference shape data. This method is based on an algorithm for obtaining conversion information, and there are various derivation methods such as a method of changing the ratio of shortening the distance according to the distance between the surfaces.

  Therefore, as described above, in the shape where it is difficult to obtain an appropriate correspondence between the shape data, such as shape data representing a planar three-dimensional shape with less unevenness, the ICP method can be used for alignment. There are many cases of failure.

  In addition, in the shape data acquired by the 3D digitizer 2, the existence direction of the 3D digitizer 2 with respect to the acquired shape data can be determined, and therefore, the correspondence relationship between the reference shape data and the reference shape data in the ICP method is acquired. Is normally performed based on the three-dimensional shape of each shape data observed only from the three-dimensional digitizer 2 side. For this reason, erroneous alignment due to erroneous recognition of the front and back sides is prevented.

  In addition, the ICP method is generally suitable for shape data with unevenness, without being affected by the positional relationship between the shape data, such as the distance between the three-dimensional shapes represented by the shape data. In many cases, the alignment can be successful regardless of the initial arrangement relationship between the shape data. This state will be described with reference to FIGS.

  FIG. 10 is a diagram illustrating an example of a different arrangement relationship between two uneven shape data. Here, the individual shape data d2a and d2b shown in FIG. 9 are adopted as the shape data having unevenness.

  The individual shape data d2a and d2b are set as standard shape data serving as a basis for alignment and reference shape data to be aligned, respectively. Further, the shape data e1 and e2 are generated by shifting the arrangement of the individual shape data d2b along the arrows in the drawing, and the shape data d2b, e1 and e2 represent the same shape, and the individual shape The data d2a has different arrangement relations.

  FIG. 11 shows that individual shape data d2b, e1, and e2 shown in FIG. 10 are coordinate-converted as shape data d2b ′, e1 ′, and e2 ′, respectively, as a result of an attempt to align them by the ICP method. The state where it is appropriately aligned with d2a is shown.

  As described above, between uneven shape data, it can be properly arranged by the ICP method with almost no influence of the original arrangement relationship, and each arranged shape data is in the initial arrangement. Regardless, the coordinates are usually converted to substantially the same arrangement.

  On the other hand, when the ICP method is applied between shape data with less unevenness, it is difficult to appropriately obtain the correspondence between the shape data. Further, if the initial arrangement relationship between the data is changed, Since the acquisition of the correspondence relationship is also affected, the correspondence relationship with the wrong correspondence relationship is more likely to be set as the initial layout deviation between the data increases.

  For this reason, when a plurality of different initial arrangement relationships are set between shape data with few irregularities and an attempt is made to align each IC using the ICP method for each arrangement relationship, the alignment usually only fails for each arrangement relationship. In addition, the arrangement of the shape data after the arrangement is varied for each arrangement relationship. Further, the variation becomes larger as the initial dislocation becomes larger. Here, a failure example of the alignment described above will be described with reference to FIG.

  FIG. 12 is a diagram illustrating an example in which alignment of shape data representing a three-dimensional shape having an arc-shaped cross section with less unevenness has failed. In FIG. 12, there are individual shape data d2f and d2g representing a three-dimensional shape having an arcuate cross section, which are set as reference shape data serving as a basis for alignment and reference shape data to be aligned, respectively. From the individual shape data d2g, shape data e3 and e4 whose positions are shifted are generated. As a result of each shape data d2g, e3 and e4 being tried to be aligned with the individual shape data d2f, the ICP method is used. Each of the shape data d2g ′, e3 ′, and e4 ′ is shown in FIG.

  In the example shown in FIG. 12, the shape data d2g ′, e3 ′, and e4 ′ have failed to be aligned with respect to the individual shape data d2f, and the respective positions vary from one another.

  In the above description referring to FIG. 10 to FIG. 12, the description is based on the premise that the shapes of the shape data are all the same and only the arrangement is different, but the shapes of the shape data are slightly different from each other. Even in the case of alignment, since alignment errors are distributed over substantially the same shape portion between the shape data to be aligned, each of the shape data with unevenness and between the shape data with few unevennesses. The result of the alignment for is usually the same as the result of the alignment described above.

  As described above, in the alignment by the ICP method between the shape data having the uneven three-dimensional shape, even if the mutual arrangement relationship is variously different, regardless of the presence or absence of some shape distortion, Usually, the arrangement is almost the same regardless of the arrangement relationship, but in the arrangement by the ICP method between the shape data having a three-dimensional shape with few irregularities, there are various mutual arrangement relationships. If they are different, the alignment usually fails regardless of the presence or absence of some shape distortion, and the arrangement of the reference shape data to be aligned after each alignment varies, and the variation The direction becomes larger as the displacement of the initial arrangement becomes larger.

  Here, for example, as described above, in the example shown in FIG. 1, although the alignment has failed, the distance between the shapes represented by the shape data is almost eliminated. Even if the success or failure (possibility) of the alignment is determined by paying attention to the mutual relationship between the shape data and the reference shape data to be aligned, it cannot always be determined correctly.

  However, for example, even in the example shown in FIG. 1, if the arrangement relationship between the standard shape data and the reference shape data is changed variously and the arrangement is tried for each arrangement relationship, it corresponds to each arrangement relationship. As described above, the mutual relationship between the arrangements of the reference shape data is usually varied.

  In the alignment feasibility determination processing according to the first embodiment, paying attention to this variation, even if the correct arrangement relationship between the reference shape data and the reference shape data is unknown, there are various initial arrangements for the reference shape data. When each of a plurality of different reference shape data is subjected to alignment with the standard shape data, if there is little variation in the shape arrangement for the plurality of reference shape data, alignment to the correct arrangement is performed. This is a process for determining whether or not alignment is possible.

  Specifically, after creating a plurality of shape data each having a different initial shape arrangement from the reference shape data with respect to the reference shape data, and trying to align each shape data to the reference shape data, The evaluation value indicating the mutual relationship of the shape arrangement with respect to the shape data is obtained, and whether or not the alignment is possible is determined by comparing the evaluation value with a determination threshold value for determining the success or failure of the alignment.

  If it is determined that the reference shape data can be aligned with the standard shape data, the device layout setting for the three-dimensional digitizer 2 that has obtained the reference shape data is also appropriate. If it is determined that the alignment is not possible, the device arrangement setting for the three-dimensional digitizer 2 is not appropriate.

  Note that, as described above, the variation in the arrangement in the interrelationship and the like usually increases as the arrangement deviation between the standard shape data and the reference shape data increases. Is appropriately adjusted according to the demands on the system.

  The details of the determination method in the alignment possibility determination process will be described later in the description of the three-dimensional shape processing apparatus 100A.

[Three-dimensional shape processing apparatus 100A:]
FIG. 3 is a block diagram illustrating the configuration of each functional unit of the three-dimensional shape data processing apparatus 100A according to the embodiment. In FIG. 3, main information exchanged between the functional units is described together with the functional units. In addition, descriptions of control signals and the like are omitted.

  As illustrated in FIG. 3, the three-dimensional shape processing apparatus 100A includes an operation unit 4, a display unit 5, an input / output unit 6, an extraction unit 7, a shape data acquisition unit 10A, a displacement distortion imparting unit 11, and a shape arrangement adjustment unit 12. The arrangement determination unit 13A, the storage unit 14, and the control processing unit 15 are mainly provided.

  When the three-dimensional shape processing apparatus 100A functions as a component of the three-dimensional shape data processing system 300A, the three-dimensional shape processing apparatus 100A determines whether or not the supplied plurality of determination shape data d1 can be appropriately arranged. An apparatus used when the 3D shape measurement system 500A measures a plurality of individual shape data using apparatus arrangement information corresponding to the shape data for determination d1 that is determined to be able to be arranged by performing the arrangement adequacy determination process to be performed. A process for obtaining the second apparatus arrangement information as arrangement information is performed.

  When the 3D shape processing apparatus 100A functions as a component of the 3D shape measurement system 500A, the 3D shape processing apparatus 100A receives a plurality of individual shape data d2 supplied from the 3D digitizer 2 as follows. Arrangement is performed using an ICP method or the like, and processing for generating shape data representing the overall three-dimensional shape of the object 1 is performed.

  The operation unit 4 shown in FIG. 3 includes, for example, a keyboard, a mouse, an operation button, and the like, and inputs an operation signal to the control processing unit 15 and an input of object identification information such as a part number. , Setting of processing apparatus parameters used for control and processing of each part of the three-dimensional shape processing apparatus 100A, object arrangement information d20 indicating the position and orientation of the object with respect to the pedestal part of the manipulator 3a, three-dimensional at the time of three-dimensional measurement Processes used and set for applications such as setting of measurement parameters d3 such as a measurement distance indicating the distance between the digitizer 2 and the object and an angle of view indicating a measurement range, and setting of device arrangement information d4 of the three-dimensional digitizer 2 Device parameters, measurement parameters d3, device arrangement information d4, and the like are stored in the storage unit 14 by the control processing unit 15. The stored measurement parameter d3 and device arrangement information d4 are acquired by the control processing unit 15 when necessary, and are supplied to each unit of the three-dimensional shape processing device 100A, the three-dimensional digitizer 2, and the manipulator control unit 3b, respectively. Examples of the processing apparatus parameter include a determination threshold th1 used for an alignment determination process described later.

  The display unit 5 is composed of, for example, a liquid crystal display, a CRT, and the like, and can display a moving image. Image information such as a three-dimensional shape of measured shape data, various messages, and an alignment possibility determination process described later. The determination result d21 (FIG. 3) or the like is displayed on the display unit 5. Further, when various parameters, device arrangement information, and the like are input from the operation unit 4, these information are displayed on the display unit 5 and used for confirmation of input values.

  The input / output unit 6 includes, for example, a multimedia drive, a network interface, and the like. The input / output unit 6 receives a storage medium 9 such as an optical disk, and exchanges information with the control processing unit 15 via the storage medium 9 and the network. Is what you do. The determination shape data d1 and individual shape data d2 from the three-dimensional digitizer 2 and the CAD data d5 from the CAD system 8 are not only directly input from the three-dimensional digitizer 2 and the CAD system 8, respectively, but also the input / output unit 6 Thus, it can also be input via the storage medium 9 or via a network.

  The storage unit 14 includes, for example, a hard disk, a ROM, a RAM, and the like. Information output from each unit of the 3D shape processing apparatus 100A such as various arranged shape data and image data, the 3D shape processing apparatus Necessary for each part of the three-dimensional shape measurement system 500A and the three-dimensional shape data processing system 300A such as processing device parameters, object placement information d20, measurement parameters d3, and device placement information d4 set to control each part of 100A. It is used for permanent storage of parameters and a control program for controlling the operation of each part of the three-dimensional shape processing apparatus 100A, and temporary storage of various information. If the 3D digitizer 2 is of a type that causes the 3D shape processing apparatus 100A to process the conversion of the measured 2D image of the object into shape data, the 3D conversion necessary for the conversion is performed. Parameters and the like are also stored in the storage unit 14.

  When the three-dimensional shape data processing system 300A is functioning, the arrangement adjusted reference shape data d16 and the reference shape data d6 obtained in the arrangement adjustment availability determination processing are stored in the storage unit 14, and then the next arrangement is made. The shape data for determination d1 to be used as the reference shape data d6 in the alignment determination process is supplied to the shape data acquisition unit 10A via the control processing unit 15.

  When the three-dimensional shape measurement system 500A is functioning, the arranged individual shape data d17 that is the arranged individual shape data d2 is used as the individual shape data to be used as the reference shape data d6 in the next arrangement. The data d2 is supplied to the shape data acquisition unit 10A via the control processing unit 15.

  The extraction unit 7, the shape data acquisition unit 10 </ b> A, the misalignment imparting unit 11, the shape arrangement adjustment unit 12, the alignment availability determination unit 13 </ b> A, and the control processing unit 15 described below execute a predetermined control program on the CPU. It may be realized by using a dedicated hardware circuit.

○ Control processing unit 15:
The control processing unit 15 controls the other functional units of the 3D shape processing apparatus 100A and exchanges necessary information with the other functional units, and also controls the 3D shape measurement system 500A and the 3D shape data processing system 300A. It controls the three-dimensional digitizer 2, the manipulator control unit 3b, and the CAD system 8, and exchanges information with them.

Shape data acquisition unit 10A:
○ Shape data acquisition unit 10A when functioning 3D shape data processing system 300A:
When the three-dimensional shape data processing system 300A is functioning, the shape data acquisition unit 10A includes a shape of the object 1 that is three-dimensionally measured from the three-dimensional digitizer 2, the extraction unit 7, and the control processing unit 15, respectively. Data, CAD data d5 about the object 1 supplied from the CAD system 8, and reference adjusted shape data d16 and reference shape data d6 corresponding to the original reference shape data d7 stored in the storage unit 14. A plurality of pieces of determination shape data d1, each having one or more pieces of determination shape data, and one or more other pieces of shape data representing substantially the same three-dimensional shape as part or all of the three-dimensional shapes represented by the shape data A plurality of determination shape data d1 including the determination shape data is appropriately supplied.

  The shape data acquisition unit 10A acquires the reference shape data d6 and the original reference shape data d7 from the supplied plurality of determination shape data d1 and supplies the reference shape data d7 to the shape arrangement adjustment unit 12, and the original reference shape data d7 is misaligned. It supplies to the grant part 11.

  From the viewpoint of improving the reliability of the alignment possibility determination process described later, it is desirable that the determination shape data has the same shape as possible as the shape data to be measured by the three-dimensional digitizer 2.

  However, in an alignment method based on shape correspondence points such as the ICP method, an overlapping portion between shapes represented by shape data is mainly used, and shape data constituting an outer edge portion of the shape is based on shape correspondence points. The shape data to be measured by the three-dimensional digitizer 2 and the shape data for determination corresponding to the shape data are slightly different from each other because they are not often used in the alignment method. However, the usefulness of the present invention is not impaired.

  FIG. 13 is a diagram illustrating the manner in which the shape data acquisition unit 10A acquires the reference shape data d6a and the original reference shape data d7a from the plurality of determination shape data d1.

  In FIG. 13A, the shape data obtained by determining the three-dimensional shape of the object 1 by the three-dimensional digitizer 2 having the measurement range n1 determined in accordance with the device arrangement information d4a is formed as the determination shape data d1a. It is supplied to the data acquisition unit 10A and acquired as reference shape data d6a by the shape data acquisition unit 10A.

  Further, in FIG. 13B, shape data obtained by determining the three-dimensional shape of the object 1 by the three-dimensional digitizer 2 having a measurement range n2 and determined in correspondence with the device arrangement information d4b is the shape data for determination d1b. Is supplied to the shape data acquisition unit 10A and acquired as original reference shape data d7a by the shape data acquisition unit 10A.

  The reference shape data d6a is shape data serving as a reference for alignment. For example, shape data for determination corresponding to shape data to be obtained in the first three-dimensional measurement among a series of three-dimensional measurements for obtaining shape data for shape inspection is obtained based on the three-dimensional measurement or CAD data Those are adopted as reference shape data. Further, the original reference shape data d7 that has already been aligned with respect to the standard shape data d6a and stored in the storage unit 14 is adopted as standard shape data for new alignment.

  The reference shape data may be set by the measurer through input from the operation unit 4 from among a plurality of pieces of determination shape data displayed on the display unit 5, or as described above, for the first three-dimensional measurement. Corresponding shape data for determination and original reference shape data already arranged may be automatically set as reference shape data. Further, when each shape data for determination is generated, the reference shape data may be set and the determination data may be added, and the additional data may be set by acquiring the shape data acquisition unit 10A.

  As the original reference shape data d7a, determination shape data that has not yet been determined whether or not it can be aligned with the standard shape data d6a is employed. For example, after the reference shape data is set, determination shape data supplied to the shape data acquisition unit 10A based on the three-dimensional measurement or CAD data is set as the original reference shape data d7a.

  14 and 15 are diagrams illustrating the standard shape data and the original reference shape data. 14 and 15, n1 to n4 are measurement ranges of the three-dimensional digitizer 2 set on a planar object (not shown), d1e to d1h are shape data for determination, and d6e to d6g are reference shape data. , D7e to d7h respectively indicate the original reference shape data. In the present application, the shape data is shown not by coordinate values but by schematically displaying the three-dimensional shape represented by the shape data. Here, for example, in the notation of n1 (d1e) (d6e) in FIGS. 14 and 15, the shape data for determination d1e corresponding to the measurement range n1 is supplied to the shape data acquisition unit 10A, and the shape data acquisition unit 10A It is shown that it is acquired as reference shape data d6e.

  FIG. 14A shows an example in which the reference shape data d6e and the original reference shape data d7f share substantially the same partial three-dimensional shape by partially overlapping the measurement ranges n1 and n2.

  FIG. 14B shows an example in which all the three-dimensional shapes represented by the standard shape data d6e are included in the three-dimensional shapes represented by the original reference shape data d7f. Such a relationship is employed, for example, when the shape of the peripheral portion that has been three-dimensionally measured at a low resolution is aligned with respect to the portion that has been previously three-dimensionally measured at a high resolution.

  FIG. 14C shows an example in which all of the three-dimensional shape represented by the original reference shape data d7f is included in the three-dimensional shape represented by the standard shape data d6e, contrary to FIG. 14B. Such a relationship is used when, for example, a large area that has been previously measured three-dimensionally at a low resolution is used as a reference, and a feature portion in that area is three-dimensionally measured at a high resolution.

  FIG. 14D shows an example in which a hole SH1 generated by a blind spot or the like is present in the three-dimensional shape represented by the reference shape data d6e shown in FIG. Such a relationship is adopted when the hole is re-measured from the direction different from the standard shape data d6e with the original reference shape data d7f so as not to cause a blind spot with higher resolution.

  FIG. 15A shows an example in which one original reference shape data d7h is acquired for the measurement leakage region SH2 with respect to the three pieces of standard shape data d6e to d6g forming the U-shaped measurement leakage region SH2. Show.

  FIG. 15B shows an example in which original reference shape data d7g corresponding to a measurement range that completely covers the two reference shape data d6e and d6f is acquired.

  Further, FIG. 15C shows that the reference shape data d6e and d6f corresponding to two partially overlapping measurement ranges n1 and n2 partially overlap with the measurement ranges n1 and n2, respectively. An example is shown in which original reference shape data d7g and d7h corresponding to partially overlapping measurement ranges n3 and n4 are acquired.

  When a plurality of reference shape data d6 (FIG. 3) is employed, the mutual relationship between the plurality of reference shape data d6 is not changed in the adjustment of the shape arrangement performed by the shape arrangement adjustment unit 12 described later. The shape arrangement with the original reference shape data d7 (FIG. 3) is adjusted.

  In addition, the mutual relationship between the shape data for determination that has been aligned in the alignment determination process performed in the past and the reference shape data d6 in the alignment determination process that is currently performed is generally not changed. Then, an alignment determination process is performed.

  Further, when a plurality of original reference shape data d7 is adopted, the alignment possibility determination process may be performed for each original reference shape data d7, or the alignment possibility determination process is simultaneously performed on the plurality of original reference shape data d7. To determine whether or not alignment is possible for each original reference shape data d7.

  As described above, the shape data acquisition unit 10A illustrated in FIG. 3 includes one or more pieces of determination shape data d1 among the plurality of pieces of determination shape data d1 each representing a plurality of three-dimensional shapes of the object 1. Is obtained as one or more reference shape data d6, and other determination shape data d1 having shape data representing a three-dimensional shape substantially the same as a part or all of the three-dimensional shape represented by the one or more reference shape data d6. Is obtained as original reference shape data d7.

  Further, the standard shape data d6 and the original reference shape data d7 represent the three-dimensional shape of the object 1 for each measurement range having portions that overlap each other on the object 1.

○ Shape data acquisition unit 10A when the three-dimensional shape measurement system 500A functions:
When the three-dimensional shape measurement system 500A is functioning, the shape data acquisition unit 10A includes a shape in which the three-dimensional digitizer 2 measures the three-dimensional shape of the object 1 from the three-dimensional digitizer 2 and the control processing unit 15, respectively. Data and a plurality of individual shape data d2 based on the arranged individual shape data d17 stored in the storage unit 14 are appropriately supplied.

  The shape data acquisition unit 10A acquires the reference shape data d6 and the original reference shape data d7 from the supplied plurality of individual shape data d2, and supplies the acquired reference shape data d7 to the shape arrangement adjustment unit 12.

Extraction unit 7:
Next, generation of shape data for determination based on CAD data will be described. The extraction unit 7 shown in FIG. 3 is a functional unit that operates when the three-dimensional shape data processing system 300A is functioning, and is connected to the CAD system 8 via an input interface (not shown). The extraction unit 7 receives CAD data d5 from the CAD system 8, and the control processing unit 15 receives measurement parameters d3, device arrangement information d4, and object arrangement information d20.

  The CAD data d5 supplied to the extraction unit 7 is first subjected to necessary coordinate transformation based on the object arrangement information d20 so as to be equal to the arrangement of the object with respect to the pedestal part of the manipulator 3a. Shape data for determination d1 representing a three-dimensional shape substantially the same as the shape data to be measured by the three-dimensional digitizer 2 arranged based on the information d4 is extracted from the CAD data d5 obtained by coordinate conversion. This extraction process is performed based on the measurement parameter d3, the device arrangement information d4, the object arrangement information d20, and the like.

  Further, after changing the arrangement of the virtual three-dimensional digitizer 2i so that the CAD data d5 and the object overlap with each other based on the coordinate system of the CAD data d5, the shape data for determination d1 based on the CAD data d5 is extracted. You may make it do.

  FIG. 16 is a diagram for explaining how the shape data for determination d1 is generated from the CAD data d5. CAD data d5 shown in FIG. 16 represents CAD data d5d representing a planar part, CAD data d5a and d5c representing a quadrangular column part and a cylindrical part provided on the planar part, and the outer edge of CAD data d5d, respectively. It consists of CAD data d5b.

  In this way, the CAD data supplied from the CAD system 8 may be configured in units of parts that form a component, or may be one unit of the entire component. Regarding the specification of a necessary part when using CAD data for each part, for example, when coordinate conversion is performed so that the CAD data for each part has the same arrangement as the object, it is based on the measurement parameter d3 and the apparatus arrangement information d4. The CAD data of each part that falls within the measurement range of the three-dimensional digitizer 2 on the object obtained in this way may be specified by requesting the CAD system 8 from the three-dimensional shape processing apparatus 100A.

  The extraction unit 7 assumes a virtual three-dimensional digitizer 2i representing the virtual three-dimensional digitizer 2 in the coordinate system of the CAD data d5 based on the device arrangement information d4a for the three-dimensional digitizer 2, and the three-dimensional digitizer 2 A virtual camera line of sight 30 is assumed for each pixel of the image sensor provided in the camera of the virtual three-dimensional digitizer 2i based on the measurement parameter d3. Here, n3 in FIG. 16 indicates the measurement range of the virtual three-dimensional digitizer 2i.

  When the virtual camera line of sight 30 is assumed, the extraction unit 7 generates the shape data for determination d1 from the CAD data d5 by obtaining the intersection coordinates with the CAD data d5 for each virtual camera line of sight 30.

  In addition, when the whole part is finely divided and the range of the CAD data for each part is smaller than or equivalent to the measurement range of the three-dimensional digitizer 2, the extraction work using the above-described virtual camera line of sight 30 is performed. The CAD data for one or more parts is output as it is as the shape data for determination d1.

  If it is determined in the alignment possibility determination process that the original reference shape data d7 cannot be aligned, the determination shape data d1 corresponding to the original reference shape data d7 is usually changed to a different one. Then, the alignment determination process is performed again.

  Here, according to the method of generating the shape data for determination d1 from the CAD data d5, the device arrangement information obtained from the original reference shape data d7 determined that the apparatus arrangement information of the virtual three-dimensional digitizer 2i cannot be arranged. It is easy to re-acquire new determination shape data d1 to be the original reference shape data d7 by changing to device arrangement information different from the above, and for measuring the determination shape data d1, It is possible to save time and work for moving the three-dimensional digitizer 2.

  Therefore, for example, the acquisition of the new determination shape data d1 and the alignment determination process based on the new determination shape data d1 are performed in the vicinity of a portion where the acquisition of the shape data of the object is desired. Even if it is repeated until the original reference shape data d7 determined to be able to be obtained is obtained, the required time is greatly reduced compared to the case where the determination shape data d1 is obtained using the three-dimensional digitizer 2. Can do.

  Further, for example, the apparatus arrangement information d4 of the virtual three-dimensional digitizer 2i for obtaining a plurality of pieces of determination shape data d1 in advance is obtained for the object 1 based on the measurement parameter d3, CAD data d5, object arrangement information d20, and the like. The overall shape data used in the shape inspection is set so as not to cause measurement omission, and the shift amount and direction of the apparatus arrangement information d4 for obtaining the new determination shape data d1 Are set in advance, without performing the three-dimensional measurement by the three-dimensional digitizer 2 and without using the object 1, the object can be obtained only by the CAD system 8 and the three-dimensional shape processing apparatus 100A. A plurality of determination shape data d1 corresponding to the desired overall shape data for the object 1 can be obtained.

  Therefore, the device arrangement of the three-dimensional digitizer 2 for measuring the shape in the shape inspection process of the actual object based on the device arrangement information d4 of the virtual three-dimensional digitizer 2i corresponding to the obtained plurality of determination shape data d1. Information d4 can be generated.

  In determining whether or not all the determination shape data d1 corresponding to the desired overall shape data for the object 1 has been obtained, the arrangement result of the determination shape data d1 displayed on the display unit 5 Based on the above, the measurer may make the determination visually, and the CAD data d5 representing the entire object 1 and the arranged shape data for determination d1 are coordinated so as to overlap in the same coordinate system. Whether or not there is the shape data for determination d1 arranged in a predetermined neighborhood at each check location, with a plurality of check locations discretely provided for the portion corresponding to the measurement desired portion of the object 1 By determining the application of the algorithm when setting the correspondence in the alignment method based on the shape corresponding points such as the ICP method and the spin image method, It may automatically detect the completion of the operations acceptability decision process allowed.

◎ Displacement imparting unit 11:
The misalignment imparting unit 11 shown in FIG. 3 is a functional unit that operates when the three-dimensional shape data processing system 300A is functioning. The misalignment imparting unit 11 is supplied with the original reference shape data d7 from the shape data acquisition unit 10A, and is supplied with the measurement accuracy d18 of the three-dimensional digitizer 2 and the positioning accuracy d19 of the manipulator 3a from the control processing unit 15, Each of the three-dimensional shapes represented by the original reference shape data d7 is one or a plurality of mutually different displacement distortions, which is at least one of the displacement of the shape arrangement indicating the position and orientation of the original shape and the distortion of the three-dimensional shape. One or a plurality of misalignment distortion-containing reference shape data d8 each representing a three-dimensional shape is generated and supplied to the shape arrangement adjusting unit 12, and used to give shape arrangement in the generation process of the misalignment distortion containing reference shape data d8. The obtained shape arrangement shift conversion M2 is supplied to the arrangement determination unit 13A.

  As the number of misalignment-containing reference shape data d8 generated increases, the shape arrangement adjustment unit 12 described later can generate more different arrangement-adjusted reference shape data d16. Prior to the measurement, many arrangement attempts can be simulated, and the reliability of the arrangement determination process can be improved.

  Here, as described in the outline of the principle of the alignment possibility determination process, as for the displacement and distortion of the shape arrangement given to the three-dimensional shape, the larger the deviation and distortion, the alignment alignment determination process is performed. The reliability with respect to the determination of the original reference shape data d7 determined to be possible increases.

  Accordingly, when it is desired to increase the reliability of the determination that the arrangement is possible due to the required specification of the system as much as possible, for example, the manipulator 3a used when acquiring the individual shape data by taking into account the safety factor. It is desirable to add to the original reference shape data at least one of the displacement and distortion of several times the arrangement determination accuracy d19 and the measurement accuracy d18 of the three-dimensional digitizer 2 respectively. In this case, the determination threshold th1 used in the alignment possibility determination unit 13A may be used in combination with a threshold that makes it difficult to determine that the alignment is possible than usual.

  In addition, for example, the system is required to balance the improvement in reliability with respect to the determination that the determination that the alignment is possible is obtained, and the measurement work and the cost of the alignment determination process. In this case, for example, it is desirable to employ a displacement and distortion of the same degree as the placement accuracy d19 of the manipulator 3a and the measurement accuracy d18 of the three-dimensional digitizer 2, respectively.

  In this case, it is possible to determine the alignment possibility determination process in a situation closer to the actual three-dimensional measurement situation. Further, depending on the requirements of the system, an arrangement displacement and distortion smaller than the arrangement determining accuracy d19 of the manipulator 3a and the measurement accuracy d18 of the three-dimensional digitizer 2 may be employed.

  Further, the displacement of the shape arrangement may be set without being based on the arrangement determining accuracy d19 of the manipulator 3a actually used. For example, even if a displacement of an appropriate shape arrangement is given based on the arrangement accuracy of a general manipulator, the usefulness of the present invention is not impaired.

  Similarly, for the distortion applied to the three-dimensional shape, an appropriate distortion may be set without being based on the measurement accuracy d18 of the three-dimensional digitizer 2.

  In addition, for example, in the case of performing the alignment feasibility determination process as a prior teaching work in the shape inspection of many mass-produced products, in addition to the arrangement determination accuracy d19 of the manipulator 3a, the object arrangement information d20 and the actual object installation time Further, the positional deviation of the object may be set in consideration of the tolerance of the positional deviation of the object.

  In addition to the misalignment imparting unit 11 in the three-dimensional shape processing apparatus 100A, the 3D digitizer 2 and the CAD system 8 may function as misalignment imparting units, respectively. In this case, the generated misalignment distortion is provided. The contained reference shape data d8 is supplied to the shape arrangement adjusting unit 12 via a data line (not shown).

  Hereinafter, the operation of the misalignment imparting unit 11 will be described with reference to FIGS. 17 to 23, taking as an example the case of generating one misalignment-containing reference shape data d8. In addition, what is necessary is just to produce | generate similarly to the case where one misalignment content reference shape data d8 is produced | generated when generating several misalignment content reference shape data d8.

  FIG. 17 is a diagram exemplifying deviation distortion-containing reference shape data generated by applying distortion. In FIG. 17, the amount of distortion applied to the three-dimensional shape is displayed exaggerated with respect to the size of the three-dimensional shape.

  First, in FIG. 17 (a), the displacement distortion imparting unit 11 imparts distortion to the three-dimensional shape represented by the original reference shape data d7i for the cylindrical target object, whereby the displacement distortion-containing reference shape data. d8i is generated.

  Specifically, the displacement distortion imparting unit 11 performs shape distortion conversion that generates distortion in the direction of deformation from the outside to the inside of the cylinder with respect to the shape data constituting the central portion of the shape represented by the original reference shape data d7i. Shape distortion transformations M1a and M1c that generate distortion in the direction of deformation from the inside of the cylinder to the outside are applied to the shape data that forms both ends of the cylinder of the shape represented by the original reference shape data d7i, while giving M1b. By giving, deviation distortion content reference shape data d8i is generated.

  Also, in FIG. 17B, the deviation distortion applying unit 11 applies the distortion to the three-dimensional shape represented by the original reference shape data d7j for the object having unevenness, thereby including the deviation distortion-containing reference shape data. d8j is generated.

  Specifically, the deviation distortion imparting unit 11 imparts a local distortion to the shape data that forms the central portion of the shape represented by the original reference shape data d7j, thereby giving a local distortion. Deviation distortion-containing reference shape data d8j having shape data d8hj representing a typical convex shape is generated. In FIG. 17B, the original reference shape data d7j and the misaligned distortion-containing reference shape data d8j are displayed slightly shifted for convenience.

  The shape distortion transformations M1a, M1b, and M1c in FIG. 17A can be obtained as transformations that vary the three-dimensional coordinates that respectively constitute the parts of the original reference shape data d7a to which these transformations are applied.

  Also, the shape distortion transformation M1d in FIG. 17B can be obtained as a transformation that varies each three-dimensional coordinate that constitutes the part of the original reference shape data d7j to which this transformation is applied.

  Here, in the conversion that gives the displacement of the shape arrangement described later, the shape data is represented by performing the same affine transformation on all the three-dimensional coordinates constituting the shape data to which the displacement is given. While the three-dimensional shape is maintained, that is, without causing distortion, the displacement of the shape arrangement is given by rotating and translating, but for example, in the application of the distortion illustrated in FIG. The shape distortion conversions M1a and M1c and the shape distortion conversion M1c are greatly different in the direction in which the three-dimensional coordinates constituting the shape data are changed.

  Also, for each part to which each shape distortion transformation is applied, the exact same transformation may be given to each three-dimensional coordinate constituting the part, or a slightly different transformation for each of these three-dimensional coordinates. May be given.

  18 and 19, the three-dimensional digitizer 2 of the three-dimensional shape data processing system 300A functions as the misalignment imparting unit 11 to generate misalignment-containing reference shape data d8a ′ with respect to the original reference shape data d7a by three-dimensional measurement. It is a figure explaining a mode to do. The conditions required for the placement accuracy d19 of the manipulator 3a when the 3D digitizer 2 functions as the misalignment imparting unit 11 will be described later in the description of the placement alignment determination unit 13A.

  As shown in FIGS. 18 and 19, the three-dimensional digitizer 2 is arranged based on the apparatus arrangement information d4b and d4b ′, and three-dimensionally measures the object 1 using the measurement ranges n2 and n2 ′. .

  The original reference shape data d7a shown in FIG. 18 is shape data measured by the three-dimensional digitizer 2 arranged based on the device arrangement information d4b. Further, the misalignment-containing reference shape data d8a 'shown in FIG. 19 is shape data measured by the three-dimensional digitizer 2 arranged based on the device arrangement information d4b'.

  In this way, by causing the 3D digitizer 2 to function as the displacement distortion imparting unit 11, the displacement of the 3D digitizer 2 with respect to the device arrangement information d4b is set, and the displacement of the shape arrangement due to the displacement of the arrangement is originally referred to. The shift distortion-containing reference shape data d8a ′ given to the shape data d7a can be acquired.

  When the 3D digitizer 2 is to function as the misalignment imparting unit 11, in order to generate a large amount of misalignment-containing reference shape data d8a ', the 3D digitizer 2 is actually arranged based on a lot of device arrangement information. However, even if the time required for teaching based on the first object becomes longer, the time required for the subsequent object measurement will be shorter based on the results of teaching. Since three-dimensional measurement can be performed in time, the usefulness of the present invention is not impaired.

  Next, a method for generating the displacement distortion-containing reference shape data corresponding to the displacement distortion-containing reference shape data d8a ′ obtained by the three-dimensional measurement shown in FIG. 19 will be described.

  FIG. 20 is a diagram for explaining the relationship between the original reference shape data d7a described in FIGS. 18 and 19 and the displacement distortion-containing reference shape data d8a '. Here, Δ1 and Δ2 indicate the mutual displacements of the measurement ranges n2 and n2 ′, respectively.

  As shown in FIG. 20, the original reference shape data d7a and the misalignment-containing reference shape data d8a ′ are shape data representing the three-dimensional shape of substantially the same portion of the object 1, and the different portions are This is a small area corresponding to the deviations Δ1 and Δ2 of the measurement range. As described above, for example, when this different part is a part of substantially the same shape part of the standard shape data d6 and the original reference shape data d7a, the arrangement by the ICP method is hardly affected. Not give. For this reason, it is possible to generate misalignment-containing reference shape data corresponding to the misalignment-containing reference shape data d8a ′ by converting the coordinates of the original reference shape data d7a.

  FIG. 21 is a diagram for explaining fluctuations in the device arrangement of the three-dimensional digitizer 2 shown in FIG.

  The coordinate systems XYZ and X′Y′Z ′ shown in FIG. 21 are initial coordinates representing the three-dimensional coordinates of the shape data measured by the three-dimensional digitizer 2 arranged based on the device arrangement information d4b and d4b ′, respectively. It represents a local coordinate system that is a system.

  In the drawings according to the present application, the circles described in the drawings showing the three-dimensional digitizers as the marks indicating the locations corresponding to the device arrangement information of the three-dimensional digitizer 2 are the local coordinates of the three-dimensional digitizers. It is the origin of the system, and the orientation of the local coordinate system and the orientation of the coordinate system representing the orientation of the device arrangement information indicating the position and orientation of each device are the same.

  Therefore, FIG. 21 shows an example in which the device arrangement information d4b ′ is given by giving the parallel movement T1 after rotating the attitude of the device arrangement information d4b by the angle θ1 around the Z axis.

  FIG. 22 and FIG. 23 are diagrams illustrating a state in which the displacement distortion-containing reference shape data d8a (FIG. 23) to which the displacement is given is generated by calculation.

  The original reference shape data d7a and the deviation distortion-containing reference shape data d8a ′ shown in FIG. 22 are respectively acquired by three-dimensional measurement using the three-dimensional digitizer 2 and the original reference shape data d7a and the deviation distortion-containing data shown in FIG. The reference shape data d8a ′ is displayed after being transformed into one coordinate system based on the relationship between the device arrangement information d4b and d4b ′.

  This coordinate transformation is performed after the deviation distortion-containing reference shape data d8a ′ obtained by the three-dimensional measurement is rotated by an angle −θ1 around the Z axis (the Z ′ axis in FIG. 21) about the device arrangement information d4b ′. , −T1 can be obtained by translation.

  FIG. 23 shows a shift generated by applying the shape placement shift conversion M2 to the original reference shape data d7a so that the original reference shape data d7a shown in FIG. 22 overlaps the shift distortion-containing reference shape data d8a ′ as much as possible. 22 shows a state where the strain-containing reference shape data d8a is added to and displayed in FIG. In the display of FIG. 23, the misalignment-containing reference shape data d8a 'and the misalignment-containing reference shape data d8a are displayed slightly shifted for convenience in order to improve visibility.

  Here, the shape displacement shift conversion M2 for obtaining the reference distortion shape data d8a containing the distortion distortion is obtained as a conversion for rotating the original reference shape data d7a by −T1 and then rotating the angle −θ1 around the Z axis.

  As in this example, the shape placement shift conversion M2 is expressed by affine conversion expressed in the form of the expression (1) or (2).

  Here, in the generation of misalignment-distorted reference shape data d8a from the original reference shape data d7a, the shape arrangement shift conversion which is the same affine transformation is performed for each of all the three-dimensional coordinates constituting the original reference shape data d7a. It can be generated by applying M2.

  In this way, after the rotation by the angle −θ1 around the Z axis from the original arrangement, the three-dimensional digitizer 2 provided with the deviation of the arrangement to be translated by −T1 actually includes the deviation distortion obtained by the three-dimensional measurement. The reference shape data d8a including misalignment distortion corresponding to the reference shape data d8a ′ is obtained by converting the original reference shape data d7a to be measured by the three-dimensional digitizer 2 in the original arrangement into the angle − It can generate | occur | produce by the calculation which rotates (theta) 1.

  By generating misalignment distortion-containing reference shape data by calculation, it is possible to generate a large amount of misalignment distortion-containing reference shape data in a shorter time than when generating misalignment distortion-containing reference shape data by three-dimensional measurement. It becomes easy to increase the number of misalignment-containing reference shape data, and the individual shape data actually acquired based on the device arrangement information corresponding to the original reference shape data determined to be alignment-possible shape data described later. The probability of successful alignment can be increased.

  In the description according to FIGS. 18 to 23 described above, a method of generating misalignment-containing reference shape data by calculation based on the actually measured original reference shape data d7a, and a method of generating misalignment-distortion-containing reference shape data by actual measurement. However, as described with reference to FIG. 16, the original reference shape data may be generated from the CAD data, and the reference shape data including misalignment distortion may be generated by performing a coordinate conversion on the original reference shape data. . Also, the reference shape data including misalignment distortion is not coordinate conversion based on the original reference shape data, but a plurality of virtual three-dimensional digitizers in which the device arrangement is shifted are assumed according to the number of necessary misalignment reference shape data, CAD Based on the data, deviation distortion-containing reference shape data corresponding to each virtual three-dimensional digitizer may be generated.

  Further, if all of the determination shape data corresponding to the reference shape data, the original reference shape data, and the deviation distortion-containing reference shape data are generated based on the CAD data, the arrangement described later is used without using the three-dimensional digitizer 2 even once. Since the apparatus arrangement information for obtaining the individual shape data by performing the alignment determination process can be generated, the three-dimensional measurement work can be omitted and the work cost can be suppressed.

  Alternatively, some of the plurality of misalignment-containing reference shape data may be generated from CAD data, and the rest may be generated from actual measurement or from measured standard shape data. The generation time can be shortened for the reference shape data containing deviation distortion generated from the CAD data. When generating the reference shape data containing deviation distortion based on the measured original reference shape data or by actual measurement, In addition, it is possible to determine whether or not the alignment can be performed in a state in which the influence of the surface state of the object such as reflectance and roughness on the actual three-dimensional measurement is reflected.

  In addition, the displacement of the shape arrangement and the distortion of the shape may be applied singly or both to one reference distortion-containing reference shape data.

  In addition, when applying the same distortion to all reference information including misalignment distortion, for example, the distortion is first applied to the original reference shape data, and then the displacement of the arrangement is applied. It is also possible to generate deviation distortion-containing reference shape data.

  As described above, the displacement distortion imparting unit 11 is at least one of the displacement of the shape arrangement indicating the position and orientation of the three-dimensional shape represented by the shape data, and the distortion of the three-dimensional shape represented by the shape data. One or a plurality of misalignment distortions each representing a three-dimensional shape given to the original reference three-dimensional shape, which is a three-dimensional shape represented by the original reference shape data d7 shown in FIG. Reference shape data d8 is generated.

◎ Shape arrangement adjusting unit 12:
○ Shape arrangement adjusting unit 12 when functioning 3D shape data processing system 300A:
When the three-dimensional shape data processing system 300A is functioning, the shape arrangement adjustment unit 12 shown in FIG. 3 receives the reference shape data d6 and the original reference shape data d7 from the shape data acquisition unit 10A, and the deviation distortion applying unit 11 The reference shape data d8 including misalignment distortion and the device arrangement information d4 from the control processing unit 15 are respectively supplied, and a plurality of arrangement adjusted reference shape data d16 are generated based on these data, and together with the reference shape data d6 It supplies to the memory | storage part 14 in order to use for the display on the display part 5, etc. In addition, a process of supplying the plurality of shape arrangement alignment conversions M3 obtained in the process of generating the plurality of arrangement adjusted reference shape data d16 to the alignment possibility determination unit 13A is performed.

  Further, as described above, the shape data corresponding to the original reference shape data d7 determined to be able to be aligned among the plurality of alignment adjusted reference shape data d16 is newly aligned with the standard shape data d6. Used as determination shape data d1 (usually used as reference shape data d6) in the determination process.

  Below, operation | movement of the shape arrangement | positioning adjustment part 12 is demonstrated, referring FIGS. 24-28. In FIG. 24 to FIG. 28, the displacement amount between the three-dimensional shapes is exaggerated with respect to the size of the three-dimensional shape.

  FIG. 24 is a diagram illustrating a state in which the standard shape data d6a, the original reference shape data d7a, and the displacement distortion-containing reference shape data d8a and d8b are converted into one coordinate system. Here, each of these data is converted into one coordinate system by the shape arrangement adjusting unit 12 based on the apparatus arrangement information d4 corresponding to the standard shape data d6a and the original reference shape data d7a.

  Here, the deviation between the standard shape data d6a and the original reference shape data d7a is caused by an error in determining the arrangement of the manipulator 3a as described above. Further, the deviation distortion-containing reference shape data d8a and d8b are generated by the deviation distortion applying unit 11 having been given different shape arrangement shift conversions M2a and M2b to the original reference shape data d7a, respectively. The deviation between the shape data d7a and the deviation distortion-containing reference shape data d8a and d8b is due to the shape arrangement shift conversion M2a and M2b.

  In addition, in the shape arrangement shift conversions M2a and M2b here, the conversion matrices in the expressions (1) and (2) are different.

  Further, a shape distortion may be further added to the misalignment-containing reference shape data d8a and d8b by further adding a shape distortion conversion (not shown).

  FIG. 25 is a diagram exemplifying a process in which the arrangement-adjusted reference shape data d16a, d16b, and d16c (FIG. 26) are generated. In FIG. 25, the shape arrangement adjustment unit 12 makes the original reference so that the original reference shape data d7a and the reference distortion shape data d8a and d8b shown in FIG. 24 overlap with the reference shape data d6a as much as possible in one coordinate system. The figure shows a state in which adjustment of the arrangement is attempted by applying shape arrangement alignment conversions M3a, M3b, and M3c to the shape data d7a and the reference distortion shape data d8a and d8b, respectively.

  FIG. 26 is a diagram exemplifying the arrangement-adjusted reference shape data d16a, d16b, and d16c that have been successfully arranged. The reference shape data d16a, d16b, and d16c whose arrangement has been adjusted are the reference shape data d7a, the reference shape data d8a and d8b containing misalignment distortion shown in FIG. 25, respectively, and the reference shape data obtained by the shape alignment conversion M3a, M3b, and M3c The shape data is appropriately arranged in the data d6a.

  FIG. 27 is a diagram exemplifying the arrangement-adjusted reference shape data d16d, d16e, and d16f for which the alignment has failed.

  In FIG. 27, the standard shape data d6c and the original reference shape data d7c have a cylindrical three-dimensional shape with little unevenness. In addition, displacement distortion-containing reference shape data d8c and d8d are generated by performing shape arrangement shift conversions M2a and M2b on the original reference shape data d7c, respectively.

  Similarly to the example shown in FIG. 26, the original reference shape data d7c and the reference distortion shape containing reference shape data d8c and d8d are respectively subjected to the shape arrangement matching conversions M3a, M3b and M3c, so that the reference shape data d6a As a result of the attempt to align the positions, the reference shape data d16d, d16e, and d16f that have been adjusted for alignment are generated, but these do not overlap the reference shape data d6c.

  That is, here, since the alignment between the shapes with less unevenness was attempted, the alignment failed.

  Regardless of the result of the alignment, the shape alignment conversions M3a, M3b, and M3c are respectively supplied as the shape alignment conversion M3 (FIG. 3) to the alignment appropriateness determination unit 13A for use in the alignment appropriateness determination process. It is done. Also, the arrangement-adjusted reference shape data d16a, d16b, and d16c are supplied to the storage unit 14 together with the reference shape data d6 as the arrangement-adjusted reference shape data d16 regardless of the result of the arrangement, and then displayed on the display unit 5. It is displayed and used for determination of whether or not the alignment can be made visually by the measurer.

  As described above, when generating a plurality of arrangement-adjusted reference shape data d16 using the original reference shape data d7a, the displacement distortion applying unit 11 generates one or more displacement distortion-containing reference shape data d8, The shape arrangement adjustment unit 12 may try to align the one or more generated distortion distortion-containing reference shape data d8 and the original reference shape data d7 with the standard shape data d6.

  Next, the case where the shape alignment matching conversion M3a is not applied to the original reference shape data d7a, that is, the case where the alignment of the original reference shape data d7a with the standard shape data d6a is not attempted will be described.

  FIG. 28 is a diagram exemplifying a process in which the placement-adjusted reference shape data d16b and d16c are generated without using the original reference shape data d7a. In FIG. 28, misalignment distortion-containing reference shape data d8a and d8b are shape data generated by adding shape placement shift conversion M2a and M2b to the original reference shape data d7a, respectively, and shape placement shift conversion Arrangement-adjusted reference shape data d16b and d16c are generated by performing shape arrangement matching conversions M3b and M3c on M2a and M2b, respectively.

  As described above, when generating a plurality of arrangement-adjusted reference shape data d16 without using the original reference shape data d7a, the displacement distortion applying unit 11 generates a plurality of displacement distortion-containing reference shape data d8, The arrangement adjustment unit 12 may try to arrange the plurality of generated distortion distortion-containing reference shape data d8 to the standard shape data d6.

  As described above, the shape arrangement adjustment unit 12 includes the original reference shape data d7 and one or more misalignment-containing reference shape data d8 shown in FIG. 3, or a plurality of misalignment-containing reference shape data d8. The original reference shape data d7 and 1 or so that substantially the same three-dimensional shape for each three-dimensional shape to be represented is superimposed on the three-dimensional shape represented by one or more standard shape data d6 in one coordinate system. Each of the plurality of shape arrangements for each three-dimensional shape represented by the plurality of deviation distortion-containing reference shape data d8 or each of the plurality of shape arrangements for each three-dimensional shape represented by the plurality of deviation distortion-containing reference shape data d8 is adjusted. As a result, a plurality of arrangement-adjusted reference shape data 16 is generated.

○ Shape arrangement adjusting unit 12 when functioning three-dimensional shape measuring system 500A:
When the 3D shape processing apparatus 100A is a functional element of the 3D shape measurement system 500A, the shape arrangement adjusting unit 12 supplies the reference shape data d6 and the original reference shape data d7 based on the individual shape data d2. Received from the data acquisition unit 10A, the arrangement of the standard shape data d6 and the original reference shape data d7 is performed to generate the arranged individual shape data d17, which is supplied to the storage unit 14.

◎ Arrangement determination unit 13A:
3A is a functional unit that operates when the three-dimensional shape data processing system 300A is functioning. Arrangement adequacy determination unit 13A is supplied with transformation matrices representing shape arrangement shift conversion M2 and shape arrangement adjustment conversion M3 from deviation distortion imparting unit 11 and shape arrangement adjustment unit 12, respectively, and based on these conversion matrices, shapes are obtained. By obtaining an evaluation value representing the mutual relationship of the shape arrangement for the plurality of arrangement adjusted reference shape data d16 generated in the arrangement adjusting unit 12, and comparing this evaluation value with the determination threshold th1, the original reference shape data d7. Are aligned with one or more reference shape data d6 supplied from the shape data acquisition unit 10A to the shape arrangement adjustment unit 12, so that one overall three-dimensional shape of the object together with the reference shape data d6. Alignment is possible to determine whether the shape data can be adjusted so that the shape can be adjusted to represent The decision process is performed. The determination threshold th1 is stored in the storage unit 14, and is supplied from the control processing unit 15 to the alignment possibility determination unit 13A.

  The determination result d21 that determines whether or not the original reference shape data d7 is shape data that can be arranged, that is, whether or not the original reference shape data d7 can be appropriately arranged, is sent to the control processing unit 15. Is supplied and displayed on the display unit 5.

  When the determination result d21 is displayed on the display unit 5, the measurer can visually adjust the alignment while changing the display state of the reference shape data d16 and the reference shape data d6 that have been adjusted in alignment displayed on the display unit 5. Therefore, it is possible to determine whether or not the alignment is possible (success / failure) based on objective numerical values, so that it is possible to prevent an erroneous determination by visual observation and to improve work efficiency.

  Next, an example of a technique for obtaining an evaluation value representing the mutual relationship between the shape arrangements of the plurality of arrangement adjusted reference shape data d16 will be described using the shape arrangement shift conversion and the shape arrangement alignment conversion shown in FIG. .

  The shape arrangement alignment conversion M3a, the shape arrangement displacement conversion M2a and the shape arrangement alignment conversion M3b, and the shape arrangement displacement conversion M2b and the shape arrangement alignment conversion M3c shown in FIG. 25, respectively, are the arrangement-adjusted reference shape data d16a, d16b, and d16c. (FIG. 26) represents “shape arrangement variation information” for each arrangement adjusted reference shape data in the process of generating from the original reference shape data d7.

  Here, the shape arrangement alignment conversion M3a, the shape arrangement displacement conversion M2a and the combination conversion M3b * M2a of the shape arrangement alignment conversion M3b, and the shape arrangement displacement conversion M2b and the combination conversion M3c * M2b of the shape arrangement alignment conversion M3c are respectively The relative shape arrangement for the arrangement-adjusted reference shape data d16a, d16b, and d16c based on the shape arrangement for the original reference shape data d7a, that is, the interrelationship of the shape arrangement for each arrangement-adjusted reference shape data. It represents.

  Note that the symbol * used in the composite conversion represents a product of the matrix and the matrix, and the product of the conversion matrix representing each conversion when the left and right conversion of the symbol * is expressed by equation (1) is combined. There is a need for a transformation matrix that represents the transformation.

  When the alignment is successful as shown in FIG. 26, the relative shape arrangements of the arrangement-adjusted reference shape data d16a, d16b, and d16c with respect to an arbitrary shape arrangement are substantially the same, so that the conversion M3a, A plurality of transformation matrices respectively representing the synthetic transformation M3b * M2a and the synthetic transformation M3c * M2b (each referred to as “relative shape arrangement transformation”) are substantially the same transformation matrix.

  However, when the alignment fails as shown in FIG. 27, the plurality of transformation matrices respectively representing the relative shape arrangement transformation for the arrangement-adjusted reference shape data d16a, d16b, and d16c are substantially the same transformation matrix. Must not.

  Therefore, as described in the outline of the principle of the alignment possibility determination process, whether alignment of the plurality of arrangement adjusted reference shape data d16 with respect to the reference shape data d6 is successful based on the respective relative shape arrangement conversion. In other words, it is possible to determine whether the reference shape data d6 and the original reference shape data d7 can reproduce the three-dimensional shape of the object with accuracy within a predetermined allowable range, and the original reference shape data It can also be determined whether or not the device arrangement information d4 of the three-dimensional digitizer 2 that has obtained d7 is also appropriate.

  As in the example shown in FIG. 28, even in the case where the original reference shape data d7a is not attempted to be aligned with the standard shape data d6a in the shape arrangement adjusting unit 12 as in the example shown in FIG. Relative shapes with respect to the reference shape data d16b and d16c whose arrangement is adjusted based on the shape arrangement with respect to the original reference shape data d7a by the relative shape arrangement conversion given as the synthesis conversion M3b * M2a and the synthesis conversion M3c * M2b An arrangement can be represented.

  Next, an evaluation value for evaluating the mutual relationship between the shape arrangements for the plurality of arrangement adjusted reference shape data d16 expressed as relative shape arrangement conversion is obtained, and the arrangement of the plurality of arrangement adjusted reference shape data d16 is performed. A method for determining whether or not the alignment is successful will be described.

  Although there are several specific determination methods, as an example of the first determination method, for example, the deviation between the corresponding components of each conversion matrix that represents the relative shape arrangement conversion is expressed as the correlation between the arrangement information. A method is adopted in which it is obtained as an evaluation value to be expressed and compared with the determination threshold th1 and if the evaluation value is equal to or less than the determination threshold th1, it is determined that the alignment is successful, and otherwise, the alignment is determined to have failed. Can do.

  In addition, as an example of the second determination method, the difference between the maximum value and the minimum value for each corresponding component of each transformation matrix representing each relative shape arrangement transformation is used as an evaluation value representing the mutual relationship between the arrangement information. A method of determining and determining that the alignment is successful if each evaluation value is equal to or less than the determination threshold th1 can be obtained and compared with the determination threshold th1. . In addition to these methods, various methods such as a combination of the first and second determination methods may be employed.

  Further, in the transformation matrix, an element representing rotation and an element representing translation may be divided as shown in the equation (2), and an evaluation value representing an interrelationship of the arrangement information may be obtained for each, The method for obtaining the evaluation value may be changed according to the state of the object such as the surface state and size of the object and the state of the three-dimensional digitizer 2 such as the measurement distance and the angle of view.

  Further, as the determination threshold th1, a determination threshold may be set for each element of the transformation matrix. For example, different judgment thresholds may be used for an element representing rotation and an element representing translation in the transformation matrix. A threshold value for determination according to a method that employs the threshold th1, and the inspection standard for the object, the state of the object such as the surface state and size of the object, and the state of the three-dimensional digitizer 2 such as the measurement distance and the angle of view. Various setting methods can be used for setting the determination threshold th1 such as changing th1 appropriately.

In the above description example, the relative shape arrangement of the reference shape data d16 whose arrangement has been adjusted based on the shape arrangement of the reference shape data d6 has been described as an example. However, in the example of FIG. Based on the shape arrangement for the distortion-containing reference shape data d8b, the relative shape arrangement conversion representing the relative shape arrangement for the arrangement-adjusted reference shape data d16a, d16b, and d16c, respectively, is M3a * M2b ⁻¹ , And M3b * M2a * M2b ⁻¹ and M3c. M2b ⁻¹ represents the inverse transformation of M2b.

These relative shape arrangement conversions are performed on each of the relative shape arrangement conversions M2a, M3b * M2a, and M3c * M2b that give the shape arrangement of each arrangement adjusted reference shape data based on the original reference shape data d7a. This is a conversion obtained by synthesizing a relative shape arrangement conversion M2b- ¹ that gives the shape arrangement of the original reference shape data d7a with the distortion-containing reference shape data d8b as a reference.

  As described above, in the process in which the arrangement-adjusted reference shape data d16a, d16b, and d16c are generated from the original reference shape data d7, the shape arrangement adjustment, which is variation information of each shape arrangement for each arrangement-adjusted reference shape data By using each of the transformation M3a, the shape placement shift conversion M2a and the shape placement alignment transformation M3b, and the shape placement displacement transformation M2b and the shape placement matching transformation M3c, the shape placement, the device placement, and any other placement in the coordinate space are used as a reference The relative shape arrangement conversion that gives the relative shape arrangement for each arrangement adjusted reference shape data can be expressed. If the relationship between coordinates is known, the arrangement in a coordinate system different from the coordinate system in which the arrangement-adjusted reference shape data exists can be used as a reference.

  Further, the interrelationship between the shape arrangements of each arrangement adjusted reference shape data based on an arbitrary arrangement in the coordinate space is the correlation between the shape arrangements of each arrangement adjusted reference shape data based on the original reference shape data d7a. Is equal to

  As described above, the alignment determination unit 13A determines that the original reference shape data d7 is one overall three-dimensional shape of one or more standard shape data d6 and original reference shape data d7 for the object 1. As shown, the plurality of arrangement-adjusted reference shape data respectively generated from the original reference shape data d7 as to whether the shape arrangement of the original reference shape data d7 can be adjusted. Judgment processing is performed based on the mutual relationship between the shape arrangements of the plurality of three-dimensional shapes represented by d16.

  If a method for determining whether or not the original reference shape data d7 can be aligned based on the mutual relationship is used, the original reference shape data d7 is generated based on the original reference shape data d7 representing the three-dimensional shape included in the measurement range on the object. One or more of the distortion-containing reference marker position data d14 and the reference shape data d6, or a plurality of third shape data that is a plurality of distortion-containing reference marker position data d14 overlaps the measurement range of the original reference shape data d7. The reference shape data d6 and the original reference shape data are based on the mutual relationship of the shape arrangement of the plurality of arrangement adjusted reference shape data d16 after being arranged with respect to the reference shape data d6 representing the three-dimensional shape of the range. It is determined whether or not the shape of the object can be reproduced with accuracy within a predetermined allowable range by the alignment with d7.

  Therefore, depending on the mutual relationship between the standard shape data d6 and the original reference shape data d7, it is determined whether or not the standard shape data d6 and the original reference shape data d7 can reproduce the shape of the object with accuracy within a predetermined allowable range. Even if it cannot be determined, it is possible to improve the probability of correctly determining whether alignment is possible.

  Further, the alignment determination unit 13A determines this mutual relationship with respect to each alignment adjusted reference shape data d16 in a process in which each of the plurality of alignment adjusted reference shape data d16 is generated from the original reference shape data d7, for example. A technique for obtaining the information based on the variation information of the shape arrangement can be adopted. Regarding the variation information of the shape arrangement for each arrangement adjusted reference shape data d16, in the process of sequential calculation using the ICP method, before obtaining each arrangement adjusted reference shape data d16 of the final arrangement, Variation information of this shape arrangement is obtained. In other words, it is possible to obtain the variation information of the shape arrangement without determining each arrangement adjusted reference shape data d16 of the final arrangement, and determine whether or not the alignment can be performed.

  If a method for obtaining the mutual relationship based on the variation information of the shape arrangement is adopted, the mutual relation is analyzed by analyzing the shape arrangement between the three-dimensional shapes respectively represented by the plurality of generated arrangement-adjusted reference shape data d16. Since it is not necessary to obtain the mutual relationship based on the variation information of the shape arrangement in the process of generating the arrangement-adjusted reference shape data d16, an accurate mutual relationship can be obtained regardless of the shape of the original reference shape data d7. The relationship can be obtained, and the probability of correctly determining whether or not the alignment is possible can be improved.

  In addition, as a technique for obtaining the mutual relationship between the shape arrangements for each arrangement adjusted reference shape data d16 by analyzing the generated plurality of arrangement adjusted reference shape data d16 itself, for example, 3 In the case where the original reference shape data d7 has features such as local irregularities of more than three points, the three or more features of the plurality of arrangement-adjusted reference shape data d16 respectively corresponding to the three or more feature points. Based on the mutual relationship between the shapes defined by the points, a method of obtaining the mutual relationship between the shape arrangements for the plurality of three-dimensional shapes respectively represented by the plurality of arrangement adjusted reference shape data d16 may be employed.

  In this case, even if the shape placement variation information for each placement adjusted reference shape data d16 is not stored, the mutual relationship between the shape placements for each placement adjusted reference shape data d16 can be obtained. it can.

  Further, the three or more feature points are not limited to the overlapping portion of the measurement range for the standard shape data d6 and the original reference shape data d7, and can be searched for in the entire original reference shape data d7.

  Here, conditions required for the placement accuracy of the manipulator 3a when the three-dimensional digitizer 2 functions as the misalignment imparting unit 11 will be described. When the three-dimensional digitizer 2 functions as the misalignment imparting unit 11, it is necessary to obtain the shape placement shift conversion M2 based on the device placement information d4 for placing the three-dimensional digitizer 2, so that the placement accuracy of the manipulator 3a is determined. If it is bad, an error from the actual value is generated in the obtained shape arrangement shift conversion M2, and therefore the determination result 10 may be erroneous.

  Therefore, when the three-dimensional digitizer 2 functions as the misalignment imparting unit 11, a plurality of different shape placement shift conversions are generated based on only the placement accuracy of the manipulator 3a, and these shape placement shift conversions are generated. Therefore, when the evaluation value of the correlation with respect to the generated arrangement-adjusted reference shape data d16 is obtained, this evaluation value needs to be smaller than the determination threshold th1 by, for example, about 1/3 or less.

  In many cases, since this condition is satisfied, the three-dimensional digitizer 2 can function as the displacement distortion applying unit 11. In 3D measurement during teaching, this condition can be easily satisfied by using a manipulator with higher positioning accuracy than the manipulator actually used in 3D measurement during shape inspection of mass-produced products. The original digitizer 2 can function as the misalignment imparting unit 11.

  In the determination method described above, whether or not the alignment is possible is determined by paying attention to the mutual relationship of the shape arrangement of each arrangement adjusted reference shape data d16. However, each piece of apparatus arrangement information d4 corresponding to each arrangement adjusted reference shape data d16. It may be determined whether or not the alignment is possible by paying attention to.

  Specifically, for example, each device distortion information that is estimated based on each piece of device arrangement information corresponding to each arrangement adjusted reference shape data d16, that is, variation information of the shape arrangement about each arrangement adjusted reference shape data d16. Each device arrangement information for the reference shape data d8 is obtained, and whether or not the alignment of the original reference shape data d7 is possible is determined by determining whether or not the obtained variation in the device arrangement information falls within a predetermined range. May be.

  Further, the difference between each of the estimated device placement information for each misalignment-containing reference shape data d8 and each of the device placement information set when each misalignment-containing reference shape data d8 is generated, It may be determined whether or not the original reference shape data d7 can be aligned by determining whether or not it falls within a predetermined range.

[Operation Flow of Apparatus and System According to Embodiment 1]
Next, the operation flow of the three-dimensional shape measurement system 500A, the three-dimensional shape data processing system 300A, and the three-dimensional shape processing apparatus 100A will be described using the flowcharts shown in FIGS.

  FIG. 29 is a flowchart illustrating an outline of a procedure step S1 that is performed by the three-dimensional shape measurement system 500A and generates shape data representing the overall three-dimensional shape of the target object 1.

  The three-dimensional shape measurement system 500A first functions as a three-dimensional shape data processing system 300A to measure three-dimensionally a plurality of individual shape data d2 arranged as overall shape data for the object 1. A process of generating a plurality of second apparatus arrangement information for arranging the three-dimensional digitizer 2 is performed (step S110).

  FIG. 30 is a flowchart illustrating an overview of a process for generating a plurality of second apparatus arrangement information described in step S110 performed by the three-dimensional shape data processing system 300A.

  First, the three-dimensional shape data processing system 300A acquires a plurality of pieces of first apparatus arrangement information indicating the position and posture of the three-dimensional digitizer 2 with respect to the base portion of the manipulator 3a for generating the shape data for determination (step S210). ). The plurality of pieces of first apparatus arrangement information are obtained by using the object arrangement information d20 indicating the position and orientation of the object with respect to the pedestal portion of the manipulator 3a to determine the relative position of the three-dimensional digitizer 2 with respect to the object 1. It is set by the measurement person inputting using the mouse or keyboard of the operation unit 4 so that the user can determine the arrangement.

  Next, it is assumed that the three-dimensional shape data processing system 300A performs three-dimensional measurement of the object 1 using the three-dimensional digitizer 2 arranged based on the plurality of first device arrangement information (device arrangement information d4). A plurality of pieces of determination shape data d1 respectively representing three-dimensional shapes that are substantially the same as the plurality of shape data to be obtained are generated (step S220).

  Since the generation of the determination shape data d1 here is the generation of the determination shape data d1 to be initially aligned, the three-dimensional digitizer 2 arranged based on the device arrangement information d4 usually performs three-dimensional measurement. It is generated based on the shape data of the object 1 obtained in this way or the CAD data d5 output by the CAD system 8.

  The generated plurality of shape data for determination d1 is supplied to the shape data acquisition unit 10A of the three-dimensional shape processing apparatus 100A, and the alignment possibility determination process is performed by the three-dimensional shape processing apparatus 100A (step S230).

  FIG. 31 is a flowchart exemplifying an outline of the alignment possibility determination process (step S230) performed by the three-dimensional shape processing apparatus 100A.

  The shape data acquisition unit 10A acquires one or more pieces of determination shape data d1 from the supplied plurality of determination shape data d1 as reference shape data d6 serving as a reference for alignment, and these reference shape data d6 represent the shape data acquisition unit 10A. Other determination shape data d1 having shape data representing a three-dimensional shape substantially the same as a part or all of the three-dimensional shape is acquired as original reference shape data d7 (step S310).

  The original reference shape data d7 acquired by the shape data acquisition unit 10A is supplied to the displacement distortion applying unit 11, and is one or both of at least one of the displacement of the shape arrangement and the distortion of the 3D shape with respect to the 3D shape. When a plurality of different displacement distortions are applied to one or more, one or more displacement distortion-containing reference shape data d8 is generated. Further, the shape placement shift conversion M2 respectively used for imparting the displacement of the shape placement is supplied to the placement alignment determination unit 13A (step S320).

  One or more misalignment-containing reference shape data d8 generated by the misalignment imparting unit 11 is supplied to the shape arrangement adjusting unit 12, and one or more standard shape data d6 and an original reference are received from the shape data acquiring unit 10A. The shape data d7 is supplied to the shape arrangement adjustment unit 12.

  The shape arrangement adjustment unit 12 is configured so that the original reference shape data d7 and one or more deviation distortion-containing reference shape data d8 or the plurality of deviation distortion-containing reference shape data d8 become one or more standard shape data d6 in one coordinate system. By adjusting a plurality of shape arrangements for each three-dimensional shape represented by the original reference shape data d7 and the one or more deviation distortion-containing reference shape data d8 so as to be overlaid, or a plurality of deviation distortion-containing references By adjusting each of the plurality of shape arrangements for each three-dimensional shape represented by the shape data d8, a plurality of arrangement adjusted reference shape data d16 is generated.

  Further, the shape arrangement matching conversion M3 used for the arrangement adjustment is supplied to the arrangement possibility determination unit 13A, and the reference shape data d6 and the plurality of arrangement adjusted reference shape data d16 are supplied to the storage unit 14 (step S330). .

  The shape placement shift conversion M2 and the shape placement conversion M3 supplied to the placement adequacy determination unit 13A are processed by the placement adequacy determination unit 13A for a plurality of three-dimensional shapes respectively represented by the plurality of placement adjusted reference shape data d16. It is converted into an evaluation value representing the mutual relationship of the shape arrangement (step S340).

  The alignment possibility determination unit 13A compares this evaluation value with the determination threshold th1 supplied from the control processing unit 15 (step S350).

  If the evaluation value is equal to or less than the determination threshold th1 as a result of the comparison, the arrangement of the plurality of arrangement adjusted reference shape data d16 has succeeded, that is, the original reference shape data d7 is appropriately arranged with respect to the standard shape data d6. It is determined that the shape data can be aligned, and the determination result d21 is supplied to the control processing unit 15 (step S360), and the process returns to step S230 (FIG. 30).

  As a result of the comparison, if the evaluation value is larger than the determination threshold th1, the alignment of the plurality of arrangement-adjusted reference shape data d16 has failed, that is, the original reference shape data d7 is appropriately arranged with respect to the standard shape data d6. It is determined that it is not the shape data that can be aligned and the determination result d21 indicating this determination result is supplied to the control processing unit 15 (step S370), and the process returns to step S230.

  The processing returned to step S230 is moved to step S240, and the control processing unit 15 confirms the determination result d21 supplied from the alignment possibility determination unit 13A, and the original reference shape data d7 is the alignmentable shape data. Check if there is any.

  Next, when the determination result d21 determines that the original reference shape data d7 is the shape data that can be aligned, the control processing unit 15 includes one or more standard shape data d6 that are aligned, and The first device arrangement information (device arrangement information d4) corresponding to the shape data for determination d1 including the original reference shape data d7 is stored as second device arrangement information in the storage unit 14 (step S260), and the process is performed in step S270. Moved to.

  Further, when the determination result d21 determines that the original reference shape data d7 is not the shape data that can be arranged, the control processing unit 15 performs the first apparatus arrangement corresponding to the original reference shape data d7. Manipulators 3a and 3b are configured so that apparatus arrangement information obtained by shifting the information by a predetermined amount is generated as new first apparatus arrangement information, and shape data for determination corresponding to the new first apparatus arrangement information is generated. The shape determining data d1 to be the new original reference shape data d7 by controlling the arrangement determining operation and the three-dimensional measuring operation of the three-dimensional digitizer 2, or controlling the operations of the CAD system 8 and the extracting unit 7. Is generated (step S250), and the process again proceeds to step S230.

  In step S270, the measurer determines whether or not the three-dimensional shape of the target portion of the target object 1 such as the target part of the shape inspection is represented without omission by the appropriately arranged shape data for determination d1. For example, by confirming an image of each of the arranged shape data for determination d1 on the display unit 5, or the control processing unit 15 obtains each piece of determination shape data d1 that is aligned with the CAD data of the entire component. Confirm by comparing each other.

  When the three-dimensional shape of the target portion of the object 1 is not represented without omission due to each of the determination shape data d1, the determination shape data corresponding to the other first apparatus arrangement information Is generated by the control processing unit 15 by controlling the manipulator, the three-dimensional digitizer 2, the CAD system 8, and the like in the next alignment possibility determination process. Shape data for determination is generated.

  Further, in the next alignment possibility determination process, determination shape data to be one or more reference shape data d6 is generated from the alignment shape data d1 that has been aligned (step S280), and the process proceeds to step S230. And the process for determining whether or not alignment is possible is repeated.

  If the three-dimensional shape of the target portion of the target object 1 is represented without omission by each determination shape data d1, the process returns to step S110 (FIG. 29). When the process returns to step S110, device arrangement information (a plurality of second device arrangement information) about the three-dimensional digitizer 2 that can measure the three-dimensional shape of the target portion of the object 1 without omission is generated. Has been.

  The process returned to step S110 is moved to step S120, and the three-dimensional shape measurement system 500A arranges the three-dimensional digitizer 2 based on each of the plurality of second apparatus arrangement information generated in step S110. A plurality of individual shape data d2 are measured (step S130).

  The plurality of measured individual shape data d2 is supplied to the shape data acquisition unit 10A of the three-dimensional shape processing apparatus 100A, and is acquired as reference shape data d6 serving as a reference for alignment or original reference shape data d7 to be aligned. Then, it is supplied to the shape arrangement adjustment unit 12, where the arrangement of the standard shape data d6 and the original reference shape data d7 is performed in the shape arrangement adjustment unit 12, and a plurality of arranged individual shape data d17 is generated. The data is supplied to the storage unit 14 and used for display on the display unit 5.

  Through the above processing, the three-dimensional shape measurement system 500A generates shape data representing the overall three-dimensional shape of the object 1.

<About Embodiment 2:>
[Three-dimensional shape measurement system 500B and three-dimensional shape data processing system 300B:]
FIG. 32 is a block diagram illustrating a configuration example of a three-dimensional shape measurement system 500B and a three-dimensional shape data processing system 300B according to the second embodiment, which is a modification of the first embodiment.

  The three-dimensional shape processing apparatus 100B, the three-dimensional digitizer 2, the manipulator 3a, the manipulator control unit 3b, and the CAD system 8 shown in FIG. 32 constitute the main parts of the three-dimensional shape measurement system 500B and the three-dimensional shape data processing system 300B. In some cases, it functions as the three-dimensional shape measurement system 500B and in other cases it functions as the three-dimensional shape data processing system 300B.

  In the three-dimensional shape data processing system 300B, a marker is attached to the three-dimensional digitizer 2 arranged based on a plurality of first device arrangement information indicating the position and posture of the three-dimensional digitizer 2 with respect to the base portion of the manipulator 3a. If the object 1b is measured from a plurality of directions so that the measurement ranges overlap, a plurality of determination shape data respectively representing three-dimensional shapes that are substantially the same as the plurality of shape data to be obtained are generated, and those Each determination shape that is determined to be alignable by determining whether it can be properly aligned based on the marker to which the shape data for determination is attached is determined by an alignment determination process (described later). First apparatus arrangement information corresponding to each data is generated as a plurality of second apparatus arrangement information.

  The three-dimensional shape measurement system 500B generates a plurality of second device arrangement information by functioning in advance as the three-dimensional shape data processing system 300B, and then, based on the generated second device arrangement information, the marker The object 1b and the three-dimensional digitizer 2 to which the mark is attached are relatively arranged, and the alignment based on the marker is applied to a plurality of individual shape data obtained by measuring the three-dimensional shape of the object 1b from a plurality of directions. Thus, shape data representing the overall three-dimensional shape of the object 1b is generated.

  In the present embodiment, the 3D shape measurement system 500B and the 3D shape data processing system 300B are configured by the same components, but the 3D shape measurement system and the 3D shape data processing system are, for example, The three-dimensional digitizer and the manipulator having the same performance may be configured by separate components including the control device thereof. For example, for the manipulator and the control device of the three-dimensional shape data processing system 300B, the three-dimensional shape You may employ | adopt the thing whose positioning accuracy is higher than the thing of the measurement system 500B.

  Note that the 3D shape data processing system 300B performs the same teaching as the 3D shape data processing system 300A for the shape inspection of many parts having the same shape.

  Of the main parts constituting the three-dimensional shape measurement system 500B and the three-dimensional shape data processing system 300B, the three-dimensional digitizer 2, the manipulator 3a, the manipulator control unit 3b, and the CAD system 8 are replaced with the three-dimensional shape processing apparatus 100A. Except for being connected to the original shape processing apparatus 100B, the configuration and operation are the same as those of the main components constituting the corresponding three-dimensional shape measurement system 500A.

  As for the three-dimensional digitizer 2 and the CAD system 8, the target for holding the three-dimensional measurement and CAD data performed by the three-dimensional digitizer 2 and the CAD system 8, respectively, is the target 1b to which a marker is attached. However, the three-dimensional shape measurement system 500A and the three-dimensional shape data processing system 300A are different from the three-dimensional digitizer 2 and the CAD system 8.

  In addition, regarding the generation of CAD data d5 (FIG. 33) for the object 1b to which the marker is attached, for example, when the shape attachment position of a mass-produced part or the like in mass production is designated in advance, the marker attachment position in each part is designated. In addition, it is possible to generate CAD data d5 of an object 1b such as a component with a marker attached thereto.

[Overview of Principle of Arrangement Acceptability Determination Processing According to Embodiment 2]
Next, prior to the description of the three-dimensional shape processing apparatus 100B, the outline of the principle of the alignment possibility determination process according to the second embodiment will be described together with the outline of the operation of the three-dimensional shape measurement system 500B.

○ Outline of operation of the three-dimensional shape measurement system 500B:
FIG. 34 is a diagram for explaining an example of setting a measurement range when the three-dimensional shape measurement system 500B measures the three-dimensional shape of the rectangular parallelepiped object 1b with a marker attached thereto from a plurality of directions. In FIG. 34, the three-dimensional digitizer 2 arranged based on the device arrangement information d4c and d4d is three-dimensionally displayed on the planar portion 1s of the object 1b to which five markers f1 to f5 are attached from a plurality of different directions. Measuring. Further, three-dimensional measurement may be performed by combining a plurality of three-dimensional digitizers such as different three-dimensional digitizers 2a and 2b shown in FIG.

  Here, the measurement ranges n3 and n4 indicate the measurement ranges of the three-dimensional digitizer 2 corresponding to the device arrangement information d4c and the device arrangement information d4d, respectively, and represent the three-dimensional shapes corresponding to the measurement ranges n3 and n4, respectively. The data d2c and d2d are supplied to the three-dimensional shape processing apparatus 100B as a plurality of individual shape data d2 (FIG. 33). The measurement ranges n3 and n4 have overlapping portions on the object 1b, and markers f1, f2, and f3 are attached to the overlapping portions.

  The individual shape data d2c and d2d respectively corresponding to the measurement ranges n3 and n4 are arranged by the three-dimensional shape processing apparatus 100B using the arrangement method based on the markers attached to the overlapping portions of the measurement ranges n3 and n4. . Thus, when using the alignment method based on the marker, the measurement is performed so that there are three or more markers that are not on the same straight line in the overlapping portion of the measurement range for acquiring the shape data to be aligned. It is necessary to set a range. In the present invention, simply “three or more markers” means “three or more markers that are not on the same straight line”.

  As the marker, a circular seal with a color that can be measured by the three-dimensional digitizer 2 such as white, or a marker in which a black ring with a predetermined width is drawn on a circular seal such as white is adopted.

  Based on the two-dimensional image measured at the time of three-dimensional measurement, three-dimensional shape data, and the like, the pasted marker is obtained with three-dimensional coordinates of a point on the marker representatively representing the position of the marker for each marker. Here, the three-dimensional coordinates of a point on the marker representatively representing the position of the marker are referred to as “three-dimensional positions” in the present invention.

  In the alignment based on the markers, for three or more markers existing in the overlapping part of the measurement range, a point on the marker that representatively represents the position of each marker, such as the center of each marker, is adopted as an “identification point”. Based on the correspondence of the three-dimensional position, which is the three-dimensional coordinate of the identification point, the arrangement of the shape data corresponding to each measurement range is performed.

  By adopting a marker-based alignment method, alignment such as planes can be performed so that alignment is not performed correctly even when an alignment method based on shape-corresponding points such as the ICP method and spin image method is applied. Can be done appropriately.

  In addition to the center point of the marker, etc., a spherical or hemispherical member is attached to the object by magnetic force or adhesive force as the identification point, and the shape data of the member obtained by three-dimensional measurement is included in the shape data of the member. You may employ | adopt the center coordinate of this member calculated | required by fitting known shape data.

  As described above, the member that gives the identification point is not limited to a planar member, and may be a three-dimensional member. Adhesive force, magnetic force, and other various methods can be adopted as a method of attaching the member to an object. .

  FIG. 35 is a diagram for explaining the individual shape data d2c and d2d acquired corresponding to the measurement ranges n3 and n4 shown in FIG. 34, respectively.

  The shape data d2cf1, d2cf2, d2cf3 and d2cf4 in the individual shape data d2c represent the three-dimensional shapes of the markers f1, f2, f3 and f4 in the measurement range n3, respectively, and the shape data d2df1, d2df2, d2df3 and the individual shape data d2d d2df5 represents the three-dimensional shape of each of the markers f1, f2, f3, and f5 in the measurement range n4.

  The individual shape data d2c and d2d are supplied to the three-dimensional shape processing apparatus 100B and arranged. At that time, for example, the shape data acquisition unit 10B (FIG. 33) uses the individual shape data d2c as the reference shape data d6. Individual shape data d2d is acquired as original reference shape data d7.

  The acquired standard shape data d6 and original reference shape data d7 are supplied to the marker position data acquisition unit 16 (FIG. 33) and the marker arrangement adjustment unit 19 (FIG. 33).

  Here, in the present application, data constituted by three-dimensional positions of one or more markers is referred to as “marker position data”, and the three-dimensional shape represented by the marker position data, that is, the tertiary of the markers constituting the marker position data. The three-dimensional shape represented by the original position is referred to as “marker position shape”.

  The marker position data acquisition unit 16 obtains the three-dimensional positions b1 to b4 and b5 to b8 of all the markers included in each of the supplied standard shape data d6 and original reference shape data d7.

  Next, among the three-dimensional positions b1 to b4 and b5 to b8 of each marker in each shape data, the marker position shape is the most mutual among the shape data of the marker position shapes for a plurality of markers. Three-dimensional positions b1 to b3 and b5 to b7 of three or more markers close to the congruence relationship are searched for, respectively.

  Next, the three-dimensional positions b1 to b3 and b5 to b7 of the searched marker are set to three or more pasted on the portions where the measurement ranges n3 and n4 for the standard shape data d6 and the original reference shape data d7 overlap, respectively. Generated as reference marker position data d11 (FIGS. 33 and 35) and original reference marker position data d12 (FIGS. 33 and 35), which are three-dimensional positions, for the marker group composed of the markers f1, f2 and f3, and obtained a correspondence relationship It supplies to the part 17 (FIG. 33) and the marker arrangement | positioning adjustment part 19 (FIG. 33).

  In the drawings of the present application, the three-dimensional position of the marker is represented as a point in the drawing determined by the three-dimensional coordinates.

  The correspondence acquisition unit 17 generates a marker data correspondence d13 (FIG. 33), which is a correspondence between the acquired reference marker position data d11 and original reference marker position data d12, and supplies it to the marker arrangement adjustment unit 19. .

  In the example of FIG. 35, the marker data correspondence d13 is given as the marker data correspondence d13a. Specifically, the marker data correspondence d13a indicates that the three-dimensional positions b5, b6, and b7 of each marker in the original reference marker position data d12 for the original reference shape data d7 are the reference marker position data d11 for the reference shape data d6, respectively. Is given as a correspondence relationship corresponding to the three-dimensional positions b1, b2 and b3 of each marker.

  The marker arrangement adjustment unit 19 has a marker data correspondence relationship between each of the three-dimensional positions b1 to b3 of each marker in the reference marker position data d11 and each of the three-dimensional positions b5 to b7 of each marker in the original reference marker position data d12. Coordinate conversion information that matches as much as possible is obtained according to the correspondence relationship of d13 (d13a), and the reference shape data d6 and the original reference shape data d7 are aligned based on this coordinate conversion information.

  In the drawings of the present application, the marker position data is represented by the marker position shape that it represents. In FIG. 35, the reference marker position data d11 and the original reference marker position data d12 are the reference marker position shape c1 and the original reference, respectively. It is represented by the marker position shape c2.

  The reference marker position shape c1 and the original reference marker position shape c2 here are substantially congruent because they are the marker position shapes for the same markers f1, f2, and f3, although they are affected by the measurement error of the three-dimensional digitizer 2. .

  FIG. 36 is a diagram for explaining a state in which the individual shape data d2c and d2d are appropriately arranged based on the marker data correspondence d13a shown in FIG.

  In FIG. 36, based on the marker data correspondence d13a, the shape data d2df1, d2df2, and d2df3 for each of the markers f1, f2, and f3 in the individual shape data d2d acquired as the original reference shape data d7 are respectively the reference shape. The marker arrangement adjustment unit 19 obtains coordinate transformation that overlaps as much as possible with the shape data d2cf1, d2cf2, and d2cf3 for the markers f1, f2, and f3 in the individual shape data d2c acquired as the data d6.

  That is, coordinate conversion is performed such that the three-dimensional positions b1 to b3 of the markers in the individual shape data d2d and d2c for the markers f1, f2 and f3 respectively match b5 to b7 as much as possible based on the marker data correspondence d13a. It is calculated | required by the marker arrangement | positioning adjustment part 19, and the individual shape data d2d are aligned with the individual shape data d2c using this coordinate transformation.

  As described above, the individual shape data d2c and d2d for each of the overlapping measurement ranges n3 and n4 is based on the three or more markers f1 to f3 in the overlapping portion of the measurement ranges n3 and n4. 100B is aligned.

○ Types of misalignment based on markers:
Next, FIGS. 37 to 40 are diagrams for explaining an alignment error caused by the marker position shape.

  In FIG. 37, a measurement range n5 is set instead of the measurement range n2 in FIG. 37, and the individual shape data d2e corresponds to the measurement range n5. Markers f1, f3, and f6 are affixed to overlapping portions of the measurement ranges n1 and n5. The marker position data v4a and v4e for the markers f1, f3, and f6 in the individual shape data d2a and d2e represent the marker position shape c4, respectively.

  FIG. 38A shows the characteristics of the marker position shape c1 represented by the marker position data v1a and v1b for the markers f2, f3, and f4 existing in the overlapping portions of the measurement ranges n1 and n2 shown in FIG. Yes. Here, in FIG. 38, coordinate axes XYZ are provided to simplify the description. In other drawings, coordinate axes are provided as appropriate.

  FIG. 38B shows the characteristics of the marker position shape c4 represented by the marker position data v4a and v4e for the markers f1, f3, and f6 existing in the overlapping portions of the measurement ranges n1 and n5 shown in FIG. Show.

  As shown in FIG. 38A, the distance in the X-axis direction between the marker f3 and the markers f2 and f4 is a length L1. As shown in FIG. 38B, the distance in the X-axis direction between the marker f3 and the markers f1 and f6 is a length L2. Here, L1 is about several times as long as L2.

  FIG. 39A shows a three-dimensional shape represented by shape data in the vicinity of the shape data d2af3 of the marker f3 in the individual shape data d2a shown in FIGS. FIG. 39 (b) shows a three-dimensional shape obtained by cutting the three-dimensional shape shown in FIG. 39 (a) along the cutting plane line AA in FIG. 39 (a).

  As shown in FIG. 39B, the three-dimensional shape represented by the shape data d2af3 of the marker f3 in the individual shape data d2a protrudes by a length L3 in the + Y direction from the shape of the peripheral portion.

  FIG. 39C shows a three-dimensional shape represented by shape data in the vicinity of the shape data d2bf3 and d2ef3 for the marker f3 in the individual shape data d2b and d2e shown in FIGS. 37 and 37, respectively. FIG. 39 (d) shows a three-dimensional shape obtained by cutting the three-dimensional shape shown in FIG. 39 (c) along the cutting plane line BB in FIG. 39 (c).

  As shown in FIG. 39D, the three-dimensional shapes represented by the shape data d2bf3 and d2ef3 of the marker f3 in the individual shape data d2b and d2e are on the same plane as the shape of the peripheral portion.

  Note that the three-dimensional shapes represented by the shape data d2af2 and d2af4 and the d2bf2 and d2bf4 respectively for the markers f2 and f4 in the individual shape data d2a and d2b in FIG. 37 are also on the same plane as the shape of the peripheral portion, respectively. .

  Similarly, the three-dimensional shapes represented by the shape data d2af1 and d2af6 and d2ef1 and d2ef6 of the markers f1 and f6 in the individual shape data d2a and d2e in FIG. 37 are also on the same plane as the shape of the peripheral portion, respectively.

  FIG. 40A shows a shape obtained by cutting the three-dimensional shape in which the individual shape data d2a and d2b in FIG. 37 are arranged based on the markers f2, f3, and f4 along the cutting line AA in FIG. ing.

  FIG. 40B shows a three-dimensional shape in which the individual shape data d2a and d2e shown in FIG. 37 are arranged based on the markers f1, f3 and f6, and cut along the cutting line AA in FIG. Is shown.

  In FIG. 40A, the shape data d2af2, d2af3 and d2af4 for the markers f2, f3 and f4 in the individual shape data d2a are the shape data d2bf2, d2bf3 and d2bf4 for the markers f2, f3 and f4 in the individual shape data d2b. For the length L3 that is the difference in the Y-axis direction between the three-dimensional shape represented by the shape data d2af3 shown in FIG. 39B and the three-dimensional shape represented by the surrounding shape data shown in FIG. In the three-dimensional shapes represented by the individual shape data d2a and d2b, there is a deviation of the angle θ1.

  In FIG. 40B, the shape data d2af1, d2af3 and d2af6 for the markers f1, f3 and f6 in the individual shape data d2a are the shape data d2ef1, d2ef3 and the shape data for the markers f1, f3 and f6 in the individual shape data d2e. The length L3 is a difference in the Y-axis direction between the three-dimensional shape represented by the shape data d2af3 shown in FIG. 39B and the three-dimensional shape represented by the surrounding shape data. For this reason, the three-dimensional shape represented by the individual shape data d2a and d2e has a deviation of the angle θ2.

  Here, the angle θ2 is several times larger than the angle θ1. This is caused by the difference between the lengths L1 and L2 shown in FIGS. 38 (a) and 38 (b), respectively.

  As described above, even if the detection errors of the three-dimensional positions of the markers are approximately the same, depending on the difference in the marker position shape for three or more markers used for alignment, the displacement for each shape data after alignment is determined. Will be very different.

  In the alignment using the ICP method or the like, if each shape corresponding to the overlapping portion of the measurement range in the one or more standard shape data and the original reference shape data is different from an allowable limit, the alignment is shifted. May occur.

  Similarly, in the alignment based on the marker, when the marker position shapes existing in the overlapping portions of the measurement ranges in the one or more standard shape data and the original reference shape data are different from each other beyond an allowable limit, As in the above example, even if the correct marker data correspondence is acquired, there is a possibility that misalignment may occur, and in the alignment based on the marker, this error is a marker used for alignment. Further amplification may occur depending on the positional relationship between the two.

  Here, in the present application, these misalignment errors are expressed as misalignment errors caused by the marker position shape.

  In the alignment between the shape data, an error in alignment caused by the marker position shape is not allowed depending on the system requirements such as the shape inspection standard. Therefore, it is required to correctly determine these errors.

○ Outline of Principle of Arrangement Acceptability Determination Process in Embodiment 2:
Next, an outline of the principle of the alignment possibility determination process in the case of performing alignment between shape data by alignment based on markers will be described together with the difference from the alignment based on the ICP method.

  In the alignment using the ICP method, as described above, the correspondence between the substantially identical shape portions in each shape data is dynamically changed for each calculation for obtaining coordinate transformation that brings the standard shape data and the original reference shape data close to each other. In order to repeatedly perform the coordinate transformation obtained based on this correspondence between the shape data, if the initial arrangement relationship between the standard shape data and the original reference shape data is different, the shape data is substantially the same. The correspondence between the shape parts also varies.

  Therefore, in particular, when the unevenness of the three-dimensional shape represented by each shape data is small, the final alignment result may differ depending on the initial arrangement relationship between the data.

  For this reason, in the alignment adequacy determination processing in the alignment based on the ICP method, the displacement of the shape arrangement indicating the position and orientation of the three-dimensional shape and the displacement distortion that is at least one of the distortion of the three-dimensional shape are converted into the original reference shape. Deviation distortion-containing reference shape data d8 (FIG. 3) is generated by giving to the three-dimensional shape represented by the data d7 (FIG. 3), and a reference shape is generated for the generated deviation distortion-containing reference shape data d8, etc. Attempt to align data d6.

  Next, the arrangement of the plurality of arrangement-adjusted reference shape data d16 and the reference shape data d6 based on the mutual relationship of the arrangement of the plurality of arrangement-adjusted reference shape data d16 (FIG. 3) generated by the arrangement. It is determined whether or not the alignment is successful, and it is determined that the original reference shape data d7 for the plurality of arrangement-adjusted reference shape data d16 for which the alignment is successful can be aligned.

  On the other hand, in the alignment based on the markers, as described above, first, a plurality of three-dimensional positions of each marker in the standard shape data d6 (FIG. 33) and the reference shape data d7 (FIG. 33) are selected. The three-dimensional positions of three or more markers whose marker position shapes are closest to each other in the mutual relationship between the shape data of the marker position shapes of the markers are respectively searched.

  Next, a marker group composed of three or more markers pasted to a portion where the measurement ranges of the reference shape data d6 and the original reference shape data d7 are overlapped with the three-dimensional positions of the searched three or more markers. Are generated as reference marker position data d11 (FIG. 33) and original reference marker position data d12 (FIG. 33), and the three-dimensional positions of the respective markers in the reference marker position data d11 are respectively set to the original reference marker position. A marker data correspondence d13 (FIG. 33) is generated that corresponds to the three-dimensional position of each marker in the data d12.

  Next, based on this marker data correspondence d13, a coordinate transformation in which the three-dimensional positions of the respective markers match as much as possible is obtained for each of the standard shape data d6 and the original reference shape data d7, and this coordinate transformation is applied to obtain the standard. The arrangement of the shape data and the original reference shape data is performed.

  Therefore, the marker data correspondence d13 is generated with almost no influence of the initial arrangement relationship between the standard shape data d6 and the original reference shape data d7, and the alignment between the standard shape data d6 and the original reference shape data d7 is also performed. This is performed without being affected by the initial arrangement relationship between the standard shape data d6 and the original reference shape data d7.

  For this reason, in the alignment possibility determination process in the alignment based on the marker, the distortion of the three-dimensional shape is given to the marker position shape represented by the original reference marker position data d12 generated from the original reference shape data d7. Generates distortion-containing reference marker position data d14 (FIG. 33), and attempts to align the generated distortion-containing reference marker position data d14 and the like with reference marker position data d11 generated from reference shape data d6. .

  Based on the mutual relationship between the plurality of arrangement adjusted marker position data generated by this arrangement, it is determined whether or not the arrangement of the plurality of arrangement adjusted marker position data and the reference marker position data d11 is successful. It is determined that the original reference shape data d7 for the plurality of arrangement-adjusted marker position data that have been successfully aligned can be aligned.

  As described above, the displacement of the shape arrangement given to the original reference marker position data d12 does not affect the arrangement of the arrangement adjusted marker position data with respect to the original reference shape data d7. Therefore, in the present invention, the original reference marker position data Only distortion may be imparted to d12, or displacement of the shape arrangement may be imparted together with the distortion.

  The reason why the interrelationship between the plurality of arrangement-adjusted marker position data is used instead of the interrelationship between the reference marker position data d11 and each of the plurality of arrangement-adjusted marker position data is as follows. In the alignment determination process in the alignment based on the shape corresponding points such as the method, a plurality of arrangement adjusted reference shape data d16 is used instead of the mutual relationship between the standard shape data d6 and each of the plurality of arrangement adjusted reference shape data d16. This is due to the same reason why the mutual relationship is used.

  That is, as long as the marker position shape expressed by the reference marker position data d11 and the plurality of placement adjusted marker position data is not completely congruent, the correlation between the reference marker position data d11 and each of the plurality of placement adjusted marker position data. This is because it cannot be determined whether or not the arrangement of the plurality of arrangement-adjusted marker position data and the reference marker position data d11 is successful.

  Note that details of the determination method in the alignment determination process for determining the alignment based on the marker will be described later in the description of the three-dimensional shape processing apparatus 100B.

[Three-dimensional shape processing apparatus 100B:]
FIG. 33 is a block diagram illustrating the configuration of each functional unit of the three-dimensional shape data processing apparatus 100B according to the modification. In FIG. 33, main information exchanged between the functional units is described together with the functional units. In addition, descriptions of control signals and the like are omitted.

  As shown in FIG. 33, the three-dimensional shape processing apparatus 100B includes an operation unit 4, a display unit 5, an input / output unit 6, an extraction unit 7, a shape data acquisition unit 10B, a marker position data acquisition unit 16, and a correspondence relationship acquisition unit. 17, a marker distortion imparting unit 18, a marker arrangement adjusting unit 19, an arrangement availability determination unit 13 </ b> B, a storage unit 14, and a control processing unit 15.

  When the 3D shape processing apparatus 100B functions as a component of the 3D shape data processing system 300B, the 3D shape processing apparatus 100B determines whether or not the supplied plurality of determination shape data d1 can be appropriately arranged. An apparatus used when the 3D shape measurement system 500B measures a plurality of individual shape data using apparatus arrangement information corresponding to the shape data for determination d1 that is determined to be able to be arranged by performing the arrangement adequacy determination process to be performed. A process for obtaining the second apparatus arrangement information as arrangement information is performed.

  In addition, when the 3D shape processing apparatus 100B functions as a component of the 3D shape measurement system 500B, a plurality of individual shape data d2 supplied from the 3D digitizer 2 is used by an alignment method based on a marker. And processing to generate shape data representing the overall three-dimensional shape of the object 1b.

  The operation unit 4, the display unit 5, the input / output unit 6, the extraction unit 7, the storage unit 14, and the control processing unit 15 illustrated in FIG. 33 are the same as the functional elements denoted by the same reference numerals in FIG. Description is omitted.

  The extraction unit 7, the shape data acquisition unit 10 </ b> B, the marker position data acquisition unit 16, the correspondence relationship acquisition unit 17, the marker distortion imparting unit 18, the marker arrangement adjustment unit 19, the alignment availability determination unit 13 </ b> B, and the control processing unit 15 include It may be realized by executing a predetermined control program by the CPU, or may be realized by using a dedicated hardware circuit.

Shape data acquisition unit 10B:
○ Shape data acquisition unit 10B when functioning 3D shape data processing system 300B:
When the three-dimensional shape data processing system 300B is functioning, the shape data acquisition unit 10B includes a shape of the object 1b measured three-dimensionally from the three-dimensional digitizer 2, the extraction unit 7, and the control processing unit 15, respectively. Based on the data, CAD data d5 for the object 1b supplied from the CAD system 8, and the reference shape data d16 and the reference shape data d6 that have been adjusted to correspond to the original reference shape data d7 stored in the storage unit 14. A plurality of pieces of determination shape data d1, which are one or more pieces of determination shape data and a three-dimensional shape that is substantially the same as part or all of the three-dimensional shape represented by the data, and a plurality of pieces of shape data that are attached to the object 1b One or more other determination shapes each having shape data representing a three-dimensional shape including three-dimensional shapes of three or more marker groups A plurality of determination shape data d1 including a chromatography data are supplied appropriately.

  The shape data acquisition unit 10B acquires one or more reference shape data d6 and original reference shape data d7 from the supplied plurality of determination shape data d1, and uses them as a marker position data acquisition unit 16 and a marker arrangement adjustment unit 19. To supply.

  From the viewpoint of confirming a measurement omission portion when arranging a plurality of pieces of determination shape data, it is desirable that the determination shape data has the same shape as possible with the shape data to be measured by the three-dimensional digitizer 2.

  However, in the alignment based on the markers, the alignment is performed based on the shape data of the same marker group including three or more markers on the object 1b, which are included in the standard shape data d6 and the original reference shape data d7, respectively. Therefore, as long as the shape data to be measured by the three-dimensional digitizer 2 and the shape data for determination corresponding to the shape data include shape data for the same marker, the cubic data represented by these shape data respectively. Even if the original shapes are slightly different from each other, the usefulness of the present invention is not impaired.

  41 and 42 are diagrams illustrating an example of generation of the standard shape data and the original reference shape data for the object 1b to which the marker is attached.

  FIG. 41 shows a state in which the three-dimensional digitizer 2 performs three-dimensional measurement for generating the determination shape data d1c and d1d (FIG. 42). Three markers f1, f2 and f3 are affixed on the object 1b, and these markers overlap two measurement ranges n3 and n4 for the three-dimensional digitizer 2 arranged based on different device arrangement information. It is affixed to the part that is.

  FIG. 42 shows how the determination shape data d1c and d1d acquired corresponding to the measurement ranges n3 and n4 shown in FIG. 41 are acquired as the reference shape data d6b and the original reference shape data d7b, respectively. Yes.

  Here, the shape data for determination d1c includes shape data d1cf1, d1cf2, and d1cf3 representing the three-dimensional shapes of the markers f1, f2, and f3, and the shape data for determination d1d includes the markers f1, f2, and f3. The shape data d1df1, d1df2, and d1df3 representing the three-dimensional shape are included.

  The determination shape data d1c and d1d supplied from the three-dimensional digitizer 2 to the shape data acquisition unit 10B are acquired as reference shape data d6b and original reference shape data d7b by the shape data acquisition unit 10B, respectively. The standard shape data d6b and the original reference shape data d7b are supplied to the marker position data acquisition unit 16 and the marker arrangement adjustment unit 19.

  In FIG. 41 and FIG. 42, the reference shape data and the original reference shape data are generated one by one from the two pieces of determination shape data. However, the reference shape data and the original reference shape data according to the second embodiment are also shown in FIG. 14 and FIG. 15, various forms similar to the standard shape data and the original reference shape data according to the first embodiment can be taken. However, as described above, it is necessary that substantially the same shape portion of the one or more standard shape data and the original reference shape data includes shape data for the same marker group including three or more markers. is there.

  Note that shape data for three or more markers exists over a plurality of reference shape data d6 that have already been properly arranged, and also exist in the original reference shape data d7 to be arranged from now on. It may be.

  Further, the supply method of the plurality of determination shape data d1 to be the standard shape data d6 and the original reference shape data d7 in the second embodiment are the same as those in the first embodiment.

  As described above, the shape data acquisition unit 10B illustrated in FIG. 33 includes 1 of the plurality of pieces of determination shape data d1 representing the plurality of three-dimensional shapes of the object 1b to which the plurality of markers are attached. The above-described determination shape data d1 is acquired as reference shape data d6, and is a three-dimensional shape substantially the same as a part or all of the three-dimensional shape represented by the one or more reference shape data d6. Other determination shape data including shape data representing a three-dimensional shape including a three-dimensional shape of a marker group including three or more markers among a plurality of pasted markers is acquired as original reference shape data d7.

○ Shape data acquisition unit 10B when the 3D shape measurement system 500B functions:
The shape data acquisition unit 10B when the 3D shape measurement system 500B functions is the shape data when the 3D shape data processing system 300B functions, except that the individual shape data d2 is supplied instead of the shape data for determination d1. The same operation as that of the acquisition unit 10B is performed.

Marker position data acquisition unit 16:
○ Marker position data acquisition unit 16 when the three-dimensional shape data processing system 300B functions:
When the three-dimensional shape data processing system 300B functions, the marker position data acquisition unit 16 shown in FIG. 33 has substantially the same shape for each of the supplied one or more reference shape data d6 and original reference shape data d7. The shape data and the like for three or more markers present in the part are extracted, and one or more reference marker position data d11 and original reference marker position data d12 respectively indicating the three-dimensional positions of the markers are generated, and the correspondence relationship While supplying to the acquisition part 17 and the marker arrangement | positioning adjustment part 19, it supplies also to the marker distortion provision part 18 about the original reference marker position data d12.

  FIG. 43 is a diagram for explaining an operation example of the marker position data acquisition unit 16 and the correspondence relationship acquisition unit 17.

  The standard shape data d6b and the original reference shape data d7b shown in FIG. 43 are the standard shape data d6b and the original reference shape data d7b shown in FIG. 42, respectively.

  The marker position data acquisition unit 16 processes the standard shape data d6b, the original reference shape data d7b, and the like, thereby obtaining the three-dimensional positions corresponding to the markers f1, f2, and f3 shown in FIG. 41 for the standard shape data d6b. Are obtained as b1, b2 and b3, and the original reference shape data d7b is obtained as b4, b5 and b6.

  The obtained three-dimensional positions b1, b2, and b3 of the markers constitute the reference marker position data v5, and the three-dimensional positions b4, b5, and b6 of the markers constitute the original reference marker position data v6. The reference marker position data v5 and the original reference marker position data v6 correspond to the reference marker position data d11 and the original reference marker position data d12 in FIG. 33, respectively.

  Specifically, first, measurement different from the overlapping portions of the markers f1, f2, and f3 and the measurement ranges n3 and n4 in FIG. 41 by image processing such as edge extraction processing from a two-dimensional image obtained at the time of three-dimensional measurement. An image of another marker (not shown) included in the range is extracted, and a pixel position corresponding to the center of each marker is obtained from this image.

  Next, the respective markers f1, f2 and f3 and other markers not shown are determined by obtaining the three-dimensional coordinates corresponding to the obtained pixel positions from the shape data such as the standard shape data d6 and the original reference shape data d7. Three-dimensional positions for all the included markers are obtained with respect to the standard shape data d6b and the original reference shape data d7b.

  Next, among the three-dimensional positions of all the markers obtained in the standard shape data d6b and the original reference shape data d7b, the marker position shape for a plurality of markers spans between the standard shape data d6b and the original reference shape data d7b. In the correlation, the three-dimensional positions of three or more markers whose marker position shapes are closest to each other are searched.

  In the shape data acquired by the three-dimensional digitizer 2, since the direction in which the three-dimensional digitizer 2 is present can be determined with respect to the acquired shape data, a marker that is usually observed only from the three-dimensional digitizer 2 side in this search. A search for a joint relationship based on the position shape is performed. For this reason, the erroneous search resulting from the misrecognition of the front and back is prevented.

  In the example of FIG. 43, the marker position shape c5 defined by the marker three-dimensional positions b1, b2, and b3 in the standard shape data d6b, and the marker defined by the marker three-dimensional positions b4, b5, and b6 in the original reference shape data d7b. The position shape c6 is substantially congruent.

  Therefore, the marker three-dimensional positions b1, b2, and b3 and b4, b5, and b6 are searched for the standard shape data d6b and the original reference shape data d7b, respectively.

  Next, the searched three-dimensional positions b1, b2 and b3, and b4, b5 and b6 of the marker are pasted on the portions where the measurement ranges n3 and n4 for the standard shape data d6b and the original reference shape data d7b respectively overlap. The reference marker position data v5 and the original reference marker position data v6, which are three-dimensional positions, for the marker group including the three or more markers f1, f2, and f3.

  Regarding the generation of the three-dimensional position of the marker, in the CAD data or the like, usually, each marker portion is registered in advance as an independent part, so the three-dimensional position of the marker is obtained without using a two-dimensional image. be able to.

  The generated reference marker position data v5 and original reference marker position data v6 are supplied to the correspondence acquisition unit 17 and the marker arrangement adjustment unit 19 as reference marker position data d11 and original reference marker position data d12 in FIG. 33, respectively. .

  As described above, the marker position data acquisition unit 16 exists in one or more pieces of the reference shape data d6 and the original reference shape data d7 in the overlapping portion on the object 1b of the measurement range for these shape data. One or more reference marker position data d11 and original reference marker position data d12 indicating each three-dimensional position of the marker group are obtained.

○ Marker position data acquisition unit 16 when the three-dimensional shape measurement system 500B functions:
The marker position data acquisition unit 16 when the three-dimensional shape measurement system 500B is functioning corresponds to the reference marker position data d11 and the original reference marker position data d12 generated by the same operation as in the function of the three-dimensional shape data processing system 300B. Supply to the relationship acquisition unit 17.

◎ Correspondence relationship acquisition unit 17:
The correspondence acquisition unit 17 illustrated in FIG. 33 performs the same operation when the 3D shape data processing system 300B functions and when the 3D shape measurement system 500B functions. The correspondence acquisition unit 17 generates a marker data correspondence d13 that is a correspondence between the one or more reference marker position data d11 and the original reference marker position data d12 supplied from the marker position data acquisition unit 16, and adjusts the marker arrangement. It supplies to the part 19.

  In the example of FIG. 43, the marker data correspondence d13b corresponds to the marker data correspondence d13. Specifically, the marker data correspondence d13b indicates the three-dimensional positions b4, b5 and b6 of each marker in the original reference marker position data v6 for the original reference shape data d7b, and the reference marker position data v5 for the reference shape data d6b. Is given as a correspondence relationship corresponding to the three-dimensional positions b1, b2 and b3 of each marker.

  The marker data correspondence d13b can be obtained from, for example, the correspondence between the vertexes of the congruent marker position shape c5 and the marker position shape c6.

  Accordingly, in FIG. 33, the marker position data acquisition unit 16 and the correspondence relationship acquisition unit 17 are different functional units, but the marker position data acquisition unit 16 and the correspondence relationship acquisition unit 17 are realized by the same functional unit. Also good.

◎ Marker distortion applying unit 18:
The marker distortion applying unit 18 illustrated in FIG. 33 is a functional unit that operates when the three-dimensional shape data processing system 300A is functioning, and is a functional unit corresponding to the deviation distortion applying unit 11 according to the first embodiment. .

  The marker distortion imparting unit 18 is supplied with the original reference marker position data d12 from the marker position data acquisition unit 16 and the measurement accuracy d18 of the three-dimensional digitizer 2 from the control processing unit 15, and the three-dimensional representation represented by the marker position data acquisition unit 16 One or a plurality of distortion-containing reference marker position data d14 each representing one or a plurality of distortion-containing reference marker position shapes to which one or a plurality of distortions different from each other are applied are generated and supplied to the marker arrangement adjustment unit 19 To do.

  In the description of the outline of the principle of the alignment possibility determination process according to the second embodiment, the marker distortion imparting unit 18 is different from the shape arrangement adjusting unit 12 according to the first embodiment, as described, and the original reference marker position shape. Be sure to give shape distortion to In other words, the displacement of the shape arrangement does not affect the determination result of the alignment adequacy determination process, and may or may not be applied, but at least the distortion of the shape is always applied.

  FIG. 44 is a diagram for explaining a generation example of distortion-containing reference marker position data v7 and v8. In FIG. 44, the distortion amount of the marker position shape is exaggerated with respect to the size of the marker position shape.

  In FIG. 44, original reference marker position data v6 representing the original reference marker position shape c6 defined by the three-dimensional positions b4, b5, and b6 of the marker is supplied from the marker position data acquisition unit 16 to the marker distortion imparting unit 18. .

  The marker distortion applying unit 18 generates a three-dimensional position b7 of the marker by applying a geometric distortion transformation M1d to the three-dimensional position b4 of the marker to thereby generate a distortion-containing reference determined by the three-dimensional positions b5, b6, and b7 of the marker. The distortion-containing reference marker position data v7 representing the marker position shape c7 is generated, and the three-dimensional position b4 of the marker is generated by adding the shape distortion transformation M1e to the three-dimensional position b6 of the marker to generate the three-dimensional position b8 of the marker. , B5 and b8, distortion-containing reference marker position data v8 representing the distortion-containing reference marker position shape c8 is generated.

  Here, the distortion-containing reference marker position shapes c7 and c8 represented by the distortion-containing reference marker position data v7 and v8, respectively, are marker position shapes obtained by adding distortion to the original reference marker position shape c6 represented by the original reference marker position data v6. is there.

  The geometric distortion transformations M1d and M1e are the same type of transformation as the geometric distortion transformation applied to the original reference shape data d7 by the deviation distortion imparting unit 11 according to the first embodiment.

  In the example of FIG. 44, the distortion-containing reference marker position data is generated by applying the shape distortion conversion to only one three-dimensional position among the three markers of the three markers constituting the original reference marker position data v6. For example, any conversion that causes distortion of at least the shape of the original reference marker position shape c6, such as moving the three-dimensional positions of the three markers constituting the original reference marker position data v6 in different directions, respectively. Conversion may be possible.

  Further, the geometric distortion conversion to be given may be set based on the measurement accuracy d18 of the three-dimensional digitizer 2 as in the case of the geometric distortion conversion imparted by the deviation distortion imparting unit 11 according to the first embodiment. An appropriate distortion transformation may be set without being based on it.

  Similarly to the displacement distortion applying unit 11, the marker distortion applying unit 18 generates one or a plurality of distortion-containing reference marker position shapes to which a plurality of different distortions are applied. Note that when only one distortion-containing reference marker position shape is generated, a plurality of arrangements are obtained by making the original reference marker position data d12 also subject to arrangement adjustment in the arrangement adjustment in the marker arrangement adjustment unit 19. Adjusted marker position data is generated.

  As described above, the marker distortion imparting unit 18 has a three-dimensional shape represented by the original reference marker position data d12 by changing at least one of the three-dimensional positions constituting the original reference marker position data d12. One or a plurality of distortion-containing reference marker position data d14 each representing one or a plurality of distortion-containing reference marker position shapes in which one or a plurality of distortions different from each other are given to the original reference marker position shape is generated.

Marker placement adjustment unit 19:
A marker arrangement adjusting unit 19 shown in FIG. 33 is a functional unit corresponding to the shape arrangement adjusting unit 12 according to the first embodiment.

○ Marker arrangement adjusting unit 19 when the three-dimensional shape data processing system 300B functions:
The marker arrangement adjustment unit 19 in the function of the three-dimensional shape data processing system 300B receives one or more reference shape data d6 and original reference shape data d7 from the shape data acquisition unit 10B and one or more reference markers from the marker position data acquisition unit 16. Position data d11 and original reference marker position data d12, marker data correspondence d13 from the correspondence acquisition unit 17, one or more distortion-containing reference marker position data d14 from the marker distortion imparting unit 18, and tertiary from the control processing unit 15 Device arrangement information d4 about the original digitizer 2 is supplied.

  The marker arrangement adjusting unit 19 that has received these various data supplies the three-dimensional position of each marker for each of the original reference marker position data d12 and the one or more distortion-containing reference marker position data d14, and one or more ones. The original reference marker position shape represented by the original reference marker position data d12 is set to 1 or 1 so that the three-dimensional position of the marker with respect to the reference marker position data d11 matches with each other in one coordinate system as much as possible according to the marker data correspondence d13. A plurality of arrangement-adjusted marker position data is obtained by adjusting a plurality of shape arrangements indicating positions and orientations of one or a plurality of distortion-containing reference marker position shapes respectively represented by the plurality of distortion-containing reference marker position data d14. Perform the operation to generate.

  The marker arrangement adjusting unit 19 also determines that the three-dimensional position of each marker for each of the plurality of distortion-containing reference marker position data d14 and the three-dimensional position of each marker for one or more reference marker position data d11 are markers. By adjusting the plurality of shape arrangements for the plurality of distortion-containing reference marker position shapes respectively represented by the plurality of distortion-containing reference marker position data d14 so as to match as much as possible in one coordinate system according to the data correspondence d13, The operation of generating the marker position data after the arrangement adjustment is performed.

  The marker arrangement adjustment unit 19 supplies the plurality of shape arrangement alignment conversions M3 used when generating the plurality of arrangement adjusted marker position data described above to the arrangement alignment determination unit 13B and also uses the plurality of shape arrangement adjustments. A plurality of arrangement-adjusted reference shape data d16 obtained by applying the transformation M3 to the original reference shape data d7 converted into one common coordinate system with the base shape data d6 based on the device arrangement information d4, The reference shape data d6 is supplied to the storage unit 14 to be used for confirmation of arrangement by visual observation.

  The operation of the marker arrangement adjustment unit 19 will be described below with reference to FIGS. 45 and 46. 45 and 46, the amount of distortion of the marker position shape and the amount of displacement between the marker position shapes are exaggerated with respect to the size of the marker position shape.

  FIG. 45 is a diagram illustrating a process in which the marker placement adjustment unit 19 generates placement-adjusted marker position data that has been successfully placed. FIG. 46 illustrates an arrangement in which the marker placement adjustment unit 19 has failed placement. It is a figure explaining the process which produces | generates adjusted marker position data.

  In FIG. 45, the reference marker position data v5 and the original reference marker position data v6 representing the reference marker position shape c5 and the original reference marker position shape c6, respectively, are based on the corresponding device arrangement information d4 of the three-dimensional digitizer 2. The marker arrangement adjusting unit 19 converts the coordinate system into one coordinate system.

  The deviation between the reference marker position data v5 and the original reference marker position data v6 here is caused by an arrangement determination error of the manipulator 3a.

  The distortion-containing reference marker position data v7 and v8 represented by the distortion-containing reference marker position shapes c7 and c8, respectively, are generated from the original reference marker position data v6 in the marker distortion applying unit 18.

  The deviation between the original reference marker position data v6 and the distortion-containing reference marker position data v7 and v8 is caused by the shape distortion conversion given to the original reference marker position data v6 in the marker distortion applying unit 18.

  Further, the three-dimensional positions of the respective markers with respect to the original reference marker position data v6 and the distortion-containing reference marker position data v7 and v8 are determined based on the marker data correspondence d13 with respect to the original reference marker position data v6. By applying shape alignment matching conversions M3d, M3e, and M3f to the original reference marker position data v6 and the distortion-containing reference marker position data v7 and v8, respectively, so as to match the three-dimensional positions of the markers as much as possible, An adjustment is attempted.

  As a result, the placement-adjusted marker position data v6 ′, v7 ′, and v8 ′ representing the placement-adjusted marker position shapes c6 ′, c7 ′, and c8 ′ are generated, and appropriately arranged with respect to the reference marker position data v5. A match has been made.

  Here, each shape alignment conversion M3d, M3e, and M3f corresponds to marker data corresponding to each vertex of the original reference marker position data v6 and distortion-containing reference marker position data v7 and v8 to be aligned, that is, the three-dimensional position of each marker. In accordance with the relationship d13, simultaneous equations are created that correspond to the three-dimensional positions of the respective markers with respect to the reference marker position data v5 by affine transformation, and the affine represented by the expression (1) or (2) using the least square method It is generated by obtaining a transformation matrix of transformation. Therefore, these affine transformations are generated before generating the arrangement-adjusted reference shape data d16.

  FIG. 46 shows an example in which the alignment is performed in the same manner as in the example of FIG. 45, but the alignment of the distortion-containing reference marker position data v8 has failed. Such an alignment failure occurs, for example, in a three-dimensional space when the marker position shape with respect to the original reference marker position data v6 is a shape that is susceptible to distortion.

  Regardless of the result of the alignment, the shape alignment conversions M3d, M3e, and M3f are respectively supplied as the shape alignment conversion M3 (FIG. 33) to the alignment appropriateness determination unit 13B and used for the alignment appropriateness determination process. It is done.

  In addition, the shape arrangement for the original reference shape data d7 converted into one common coordinate system with the reference shape data d6 based on the device arrangement information d4 corresponding to the reference shape data d6 and the original reference shape data d7, respectively. A plurality of arrangement-adjusted reference shape data d16 obtained by applying the alignment conversions M3d, M3e, and M3f and the standard shape data d6 are transferred to the storage unit 14 for visual confirmation of the alignment of the display image. Supplied.

  In the second embodiment, one marker data correspondence d13 obtained as the correspondence between the standard marker position data d11 and the original reference marker position data d12 is used as the original reference marker position data v6 and each distortion-containing reference marker position data. Although alignment is performed by applying to all of v7 and generating a plurality of arrangement-adjusted marker position data, another method may be employed.

  As an alternative method, for example, after each distortion-containing reference marker position data v7 is generated, all the markers included in the standard shape data d6 are targeted and are most congruent with each of the distortion-containing reference marker position data v7. Each marker position data representing a near marker position shape is newly searched, and marker data correspondence is acquired for each combination of distortion-containing reference marker position data and newly searched marker position data. A technique of generating a plurality of placement-adjusted marker position data by aligning each distortion-containing reference marker position data v7 with a corresponding marker position shape based on the data correspondence relationship may be employed.

  According to this alternative method, an erroneous alignment caused by the distortion of the marker position shape caused by a measurement error or the like is simulated over a wider range than when the alignment is attempted based on one marker data correspondence d13. can do.

  The search for new marker position data in the alternative method may be performed not only on the reference shape data d6 but also on the original reference shape data d7 determined to be able to be aligned in the past.

○ Marker arrangement adjustment unit 19 when the three-dimensional shape measurement system 500B functions:
The marker arrangement adjusting unit 19 at the time of the function of the three-dimensional shape measurement system 500B receives one or more reference shape data d6 and original reference shape data d7 based on a plurality of individual shape data from the shape data acquisition unit 10B, and a marker position data acquisition unit. From 16 to one or more standard marker position data d11 and original reference marker position data d12, from the correspondence acquisition unit 17 to the marker data correspondence d13, from the control processing unit 15 to the device arrangement information d4 about the three-dimensional digitizer 2, respectively. Supplied.

  The marker arrangement adjusting unit 19 performs the 3D processing of each marker with respect to the original reference marker position data d12 in accordance with the marker data correspondence d13 obtained by the marker position data acquiring unit 16 as in the case of the function of the 3D shape data processing system 300B. Shape arrangement alignment conversion is obtained in which the position is matched with the three-dimensional position of each marker with respect to the reference marker position data d11 as much as possible.

  Next, the marker arrangement adjusting unit 19 performs the obtained shape arrangement adjustment conversion on the basis of the apparatus arrangement information d4 corresponding to the standard shape data d6 and the original reference shape data d7, respectively. By applying to the original reference shape data d7 converted to the coordinate system, the standard shape data d6 and the original reference shape data d7 are aligned.

  The aligned standard shape data d6 and original reference shape data d7 are supplied to the storage unit 14 as the aligned individual shape data d17.

◎ Arrangement determination unit 13B:
33 is a functional unit that operates when the three-dimensional shape data processing system 300B is functioning. The arrangement adequacy determination unit 13B is supplied with a conversion matrix representing the shape arrangement adjustment conversion M3 from the marker arrangement adjustment unit 19, and a plurality of arrangement adjustments generated in the marker arrangement adjustment unit 19 based on these conversion matrices. An evaluation value representing the mutual relationship of the shape arrangement with respect to the marker position data is obtained.

  The alignment adequacy determination unit 13B compares the evaluation value with the determination threshold th1 so that the original reference shape data d7 is supplied to the marker arrangement adjustment unit 19 from the shape data acquisition unit 10B. Whether or not the shape data can be arranged so that the shape arrangement can be adjusted so as to represent one overall three-dimensional shape of the object together with the reference shape data d6 by being aligned with the data d6. A determination process for determining the alignment is performed.

  Note that the determination threshold th1 is stored in the storage unit 14, and is supplied from the control processing unit 15 to the alignment possibility determination unit 13B.

  The determination result d21 that determines whether or not the original reference shape data d7 is shape data that can be arranged, that is, whether or not the original reference shape data d7 can be appropriately arranged, is sent to the control processing unit 15. Is supplied and displayed on the display unit 5.

  When the determination result d21 is displayed on the display unit 5, the measurer can visually adjust the alignment while changing the display state of the reference shape data d16 and the reference shape data d6 that have been adjusted in alignment displayed on the display unit 5. Therefore, it is possible to determine whether or not the alignment is possible (success / failure) based on objective numerical values, so that it is possible to prevent an erroneous determination by visual observation and to improve work efficiency.

  Next, an example of a technique for obtaining an evaluation value representing a mutual relationship between shape arrangements for a plurality of arrangement-adjusted marker position data using the shape arrangement alignment conversion shown in FIGS. 45 and 46 will be described.

  The shape alignment conversions M3d, M3e, and M3f shown in FIG. 45 are respectively performed in the process in which the alignment adjusted marker position data v6 ′, v7 ′, and v8 ′ are generated from the original reference marker position data v6. It represents “shape arrangement variation information” for the arrangement-adjusted marker position data.

  Here, since the change in the shape arrangement of each distortion-containing reference marker position data based on the shape arrangement of the original reference marker position data v6 is extremely small compared to the shape arrangement alignment conversions M3d, M3e, and M3f, The transformations M3d, M3e, and M3f are respectively the relative shape arrangements for the arrangement-adjusted marker position data v6 ′, v7 ′, and v8 ′ based on the shape arrangement for the original reference marker position data v6, ie, This indicates the mutual relationship between the shape arrangements for each of the marker position adjusted data.

  As shown in FIG. 45, when the alignment of each arrangement-adjusted marker position data is successful, the relative shapes of the arrangement-adjusted marker position data v6 ′, v7 ′, and v8 ′ with respect to an arbitrary shape arrangement Since the arrangement is substantially the same shape arrangement having variations within a predetermined range determined by the inspection standard or the like, a plurality of transformation matrices respectively representing the shape arrangement alignment transformations M3d, M3e, and M3f are substantially the same transformation matrix. Become.

  However, as shown in FIG. 46, when the alignment for each of the adjusted marker position data fails, the relative shape arrangement conversion for the adjusted marker position data v6 ′, v7 ′, and v8 is represented. The plurality of transformation matrices are not substantially the same transformation matrix.

  Therefore, as described in the outline of the principle of the alignment possibility determination process, is the alignment of the plurality of arrangement adjusted marker position data with respect to the reference marker position data v5 based on the respective relative shape arrangement conversions successful? In other words, it is possible to determine whether or not the original reference shape data d7 is alignable shape data that can be aligned with respect to the standard shape data d6, and includes a marker used for alignment It is also possible to determine whether or not the device arrangement information d4 of the three-dimensional digitizer 2 that has obtained the original reference shape data d7 is also appropriate.

  Even when the original reference marker position data v6 is not attempted to be aligned with the reference marker position data v5 in the marker arrangement adjustment unit 19, the shape arrangement adjustment variables M3e and M3f are similar to the example shown in FIG. Relative shape arrangement about the arrangement-adjusted marker position data v7 ′ and v8 ′ based on the shape arrangement about the original reference marker position data v6 can be expressed by the relative shape arrangement conversion given as follows.

  In addition, the evaluation value for evaluating the mutual relationship of the shape arrangement for the plurality of arrangement adjusted marker position data expressed as relative shape arrangement conversion is obtained, and the arrangement adjustment of the plurality of arrangement adjusted marker position data is successful. As for the evaluation value acquisition method and the determination threshold value th1 setting method regarding the method for determining whether or not the determination is made, various acquisition methods similar to the determination method in the description of the alignment feasibility determination unit 13A according to the first embodiment. And a setting method may be adopted.

  As described above, in the example of FIG. 45, each of the arrangement-adjusted marker position data in the process in which the arrangement-adjusted marker position data v6 ′, v7 ′, and v8 ′ are generated from the original reference marker position data v6. Relative with respect to each arrangement-adjusted marker position data with reference to any arrangement in the shape arrangement, apparatus arrangement, and other coordinate space by each of the shape arrangement alignment conversions M3d, M3e, and M3f, which are variation information of the shape arrangement Relative shape placement transformations that give a typical shape placement. If the relationship between coordinates is known, the arrangement in a coordinate system different from the coordinate system in which the arrangement-adjusted reference shape data exists can be used as a reference.

  As described above, the alignment determination unit 13B determines that the original reference shape so that the one or more standard shape data d6 and the original reference shape data d7 represent one overall three-dimensional shape of the object. Whether or not the shape arrangement of the data d7 can be adjusted based on the marker data correspondence d13 is based on the mutual relationship of the shape arrangements of the plurality of three-dimensional shapes respectively represented by the plurality of arrangement adjusted marker position data. Judgment processing is performed.

  When it is determined whether or not the two markers can be aligned based on the mutual relationship between the reference marker position data d11 and the original reference marker position data d12, if the marker position shape represented by each is a complete congruent figure, the marker data is supported. According to the relationship d13, a state in which the three-dimensional positions of the respective markers are completely matched can be determined as a correct alignment state. However, due to the measurement error of the three-dimensional digitizer 2, the reference marker position data d11 and the original reference are determined. In practice, it is almost impossible for the marker position shapes represented by the marker position data d12 to be congruent.

  Therefore, depending on the correlation between the reference marker position data d11 and the original reference marker position data d12, the reference marker position data d11 and the original reference marker position data d12 can reproduce the shape of the object with an accuracy within a predetermined allowable range. It cannot be determined whether or not.

  On the other hand, the method of the present embodiment is used to determine whether or not alignment of the original reference shape data d7 is possible based on the mutual relationship of the shape arrangement of a plurality of arrangement-adjusted marker position data representing different marker position shapes. Then, the original reference marker position data expressing the positions of three or more target identification points in the measurement range (second region) on the object provided with a plurality of identification points represented by the three-dimensional positions of the markers. A plurality of distortion-containing reference marker position data d14 generated from d12 or a plurality of third position data composed of one or more distortion-containing reference marker position data d14 and original reference marker position data d12 is original reference marker position data d12. Represents the position of each of the three or more target identification points in the measurement range (first region) having a portion overlapping with the measurement range of The three-dimensional shape of the object for the first region and the second region is generated based on the mutual relationship between the shape placements of the plurality of placement adjusted marker position data generated by placing the reference marker position data d11. It is determined whether or not each shape data expressed can reproduce the three-dimensional shape of the object with accuracy within a predetermined allowable range.

  Therefore, depending on the mutual relationship between the shape arrangements of the reference marker position data d11 and the original reference marker position data d12, each shape data representing the three-dimensional shape of the object for the first region and the second region is predetermined. Even if it is not possible to determine whether or not the three-dimensional shape of the object can be reproduced with an accuracy within the allowable range, it is possible to improve the probability of correctly determining whether or not alignment is possible.

  Further, the alignment possibility determination unit 13B determines this mutual relationship, for example, the shape of each alignment adjusted marker position data in the process in which each of the plurality of alignment adjusted marker position data is generated from the original reference marker position data d12. A technique for obtaining the information based on the variation information of the arrangement can be adopted.

  If this method is used to obtain the interrelationship based on the variation information of the shape arrangement, the interrelationship is determined by analyzing the shape arrangement between the three-dimensional shapes represented by each of the generated arrangement-adjusted marker position data. Since it is not necessary to obtain the interrelationship based on the variation information of the shape arrangement in the process of generating the arrangement adjusted marker position data, regardless of the marker position shape of the original reference marker position data d12, Accurate correlation can be obtained, and the probability of correctly determining whether or not alignment is possible can be improved.

  In addition, for example, the three-dimensional positions of the three or more markers of the plurality of arrangement-adjusted marker position data respectively corresponding to the three-dimensional positions of the three or more markers that are not on the same straight line constituting the original reference marker position data d12 are represented. Based on the mutual relationship between the marker position shapes, a method of obtaining the mutual relationship of the shape arrangements for the marker position shapes respectively represented by the plurality of arrangement-adjusted marker position data may be employed.

  In addition, similar to the alignment determination unit 13A according to the first embodiment, the alignment determination unit 13B also determines whether or not alignment is possible by paying attention to each piece of device arrangement information d4 corresponding to each arrangement adjusted marker position data. May be determined.

[Operation Flow of Apparatus and System According to Second Embodiment:]
Next, the operation flow of the three-dimensional shape measurement system 500B, the three-dimensional shape data processing system 300B, and the three-dimensional shape processing apparatus 100B will be described using the flowcharts shown in FIGS. 29, 30 and 47.

  FIG. 29 is a flowchart illustrating an outline of a procedure step S1B for generating shape data representing the overall three-dimensional shape of the object 1b, which is performed by the three-dimensional shape measurement system 500B.

  FIG. 30 is a flowchart exemplifying an outline of processing for generating a plurality of second apparatus arrangement information described in step S110 of FIG. 29 performed by the three-dimensional shape data processing system 300B.

  FIG. 47 is a flowchart exemplifying an outline of the alignment possibility determination process (step S230B in FIG. 30) performed by the three-dimensional shape processing apparatus 100B.

  The flowcharts shown in FIGS. 29, 30 and 31 according to the first embodiment and the flowcharts shown in FIGS. 29, 30 and 47 according to the second embodiment are substantially the same as those shown in FIGS. Only the flowchart of the alignment possibility determination process (steps S230 and S230B) described in 47 is different.

  Therefore, as an operation flow according to the second embodiment, only the alignment possibility determination process (step S230B) described in FIG. 47 will be described below.

  In step S410 of FIG. 47, the shape data acquisition unit 10B uses one or more pieces of determination shape data d1 as a reference shape among the plurality of determination shape data d1 for the object 1b to which the plurality of supplied markers are attached. It is acquired as data d6 and is a three-dimensional shape substantially the same as a part or all of the three-dimensional shape represented by one or more reference shape data d6, and three or more of a plurality of markers attached to the object 1b. Other determination shape data d1 including shape data representing a three-dimensional shape including the three-dimensional shape of the marker group is acquired as original reference shape data d7, and one or more standard shape data d6 and original reference shape data d7 are obtained. Is supplied to the marker position data acquisition unit 16 and the marker arrangement adjustment unit 19.

  Next, the marker position data acquisition unit 16 obtains the three-dimensional positions of all the markers included in the supplied one or more reference shape data d6 and original reference shape data d7, and all of the reference shape data d6. Among the three-dimensional positions of the markers and the three-dimensional positions of all the markers for the original reference shape data d7, in the interrelationship between these shape data, the marker position shapes of three or more markers that are closest to each other The three-dimensional position is searched, the three-dimensional positions of the three or more markers searched in the one or more reference marker position data d11 are generated as the one or more reference marker position data d11, and the correspondence acquisition unit 17 and the marker arrangement adjustment It supplies to the part 19 (step S420).

  Further, three-dimensional positions of three or more markers searched in the original reference marker position data d12 are generated as original reference marker position data d12 and supplied to the correspondence acquisition unit 17, the marker distortion imparting unit 18, and the marker arrangement adjustment unit 19. (Step S430).

  The correspondence relationship acquisition unit 17 sets each of the reference marker position data d11 based on the correspondence between the substantially congruent marker position shapes represented by each of the one or more supplied reference marker position data d11 and the original reference marker position data d12. A marker data correspondence d13 representing the correspondence between the three-dimensional position of each marker and the three-dimensional position of each marker for the original reference marker position data d12 is generated and supplied to the marker arrangement adjusting unit 19 (step S440).

  The marker distortion imparting unit 18 is based on the measurement accuracy d18 of the three-dimensional digitizer 2 separately supplied from the control processing unit 15, and at least one of the three-dimensional positions of each marker with respect to the supplied original reference marker position data d12. Is changed to one or a plurality of distortion-containing reference marker position shapes obtained by adding one or a plurality of different distortions to the original reference marker position shape which is a three-dimensional shape represented by the original reference marker position data d12. One or a plurality of distortion-containing reference marker position data d14 to be expressed is generated and supplied to the marker arrangement adjusting unit 19 (step S450).

  Based on the supplied marker data correspondence d13, the marker arrangement adjustment unit 19 includes the original reference marker position data d12 and one or more distortion-containing reference marker position data d14 or a plurality of distortion-containing reference marker position data d14. Marker represented by the original reference marker position data d12 so that the three-dimensional position of each marker for each and the three-dimensional position of each marker for one or more reference marker position data match as much as possible in one coordinate system. Adjusting a plurality of shape arrangements indicating the position and orientation of each of the position shape and the marker position shape represented by one or a plurality of distortion-containing reference marker position data d14, or a plurality of distortion-containing reference marker position data d14 By adjusting multiple shape arrangements for the marker position shape represented by Te, and generates a plurality of placement adjusted marker position data.

  In addition, the marker arrangement adjustment unit 19 supplies the shape arrangement alignment conversion M3, which is a plurality of affine transformations obtained in the generation process of the plurality of arrangement adjusted marker position data, to the arrangement adequacy determination unit 13B, and the reference shape data d6 and the reference shape data d6 arranged by the shape arrangement matching conversion M3 are supplied to the storage unit 14 as arrangement adjusted reference shape data d16 (step S460).

  The shape alignment matching conversion M3 supplied to the alignment possibility determination unit 13B represents the mutual relationship between the shape arrangements of the plurality of three-dimensional shapes respectively represented by the plurality of arrangement adjusted marker position data by the alignment possibility determination unit 13B. It is converted into an evaluation value (step S470).

  The alignment possibility determination unit 13B compares this evaluation value with the determination threshold th1 supplied from the control processing unit 15 (step S480).

  If the evaluation value is equal to or less than the determination threshold th1 as a result of the comparison, the arrangement of the plurality of arrangement-adjusted marker position data has succeeded, that is, the original reference shape data d7 is appropriately arranged with respect to the reference shape data d6. It is determined that the shape data can be arranged and the determination result d21 indicating the determination result is supplied to the control processing unit 15 (step S490), and the process returns to step S230B (FIG. 30).

  As a result of the comparison, if the evaluation value is larger than the determination threshold th1, the alignment of the plurality of arrangement adjusted marker position data has failed, that is, the original reference shape data d7 is appropriately aligned with the reference shape data d6. It is determined that the shape data is not possible to be aligned, and the determination result d21 indicating this determination result is supplied to the control processing unit 15 (step S492), and the process is returned to step S230B to determine whether the alignment is possible. The entire process ends.

  Arrangement availability determination processing by the three-dimensional shape processing apparatus 100B described above, processing for generating a plurality of second apparatus arrangement information by the three-dimensional shape data processing system 300B described in FIG. 30, and description in FIG. Through the above processing, the three-dimensional shape measurement system 500B generates shape data representing the overall three-dimensional shape of the object 1b.

500A, 500B Three-dimensional shape measurement system 300A, 300B Three-dimensional shape data processing system 100A, 100B Three-dimensional shape data processing device 1, 1a, 1b Object 2, 2a, 2b Three-dimensional digitizer 2i Virtual three-dimensional digitizer 3a Manipulator 3b Manipulator Control unit 4 Operation unit 5 Display unit 6 Input / output unit 7 Extraction unit 8 CAD system 9 Recording medium 10A, 10B Shape data acquisition unit 11 Displacement distortion imparting unit 12 Shape arrangement adjustment unit 13A, 13B Arrangement availability determination unit 14 Storage unit 15 Control processing unit 16 Marker position data acquisition unit 17 Correspondence relationship acquisition unit 18 Marker distortion imparting unit 19 Marker placement adjustment unit 30 Virtual camera line of sight n1, n2, n1 ′, n2 ′, n3, n4 Measurement range th1 determination threshold d1, d1a, For d1b, d1c, d1d judgment Shape data d2, d2a, d2b, d2c, d2d Individual shape data d3 Measurement parameters d4, d4a, d4b, d4b ′, d4c, d4d Device arrangement information d5, d5a, d5b, d5c, d5d CAD data d6, d6a, d6b Reference shape Data d7, d7a, d7b Original reference shape data d8, d8a, d8a ′, d8b Deviation distortion-containing reference shape data d11, v5 Reference marker position data d12, v6 Original reference marker position data d13, d13a, d13b Marker data correspondence d14, v7, v8 Distortion-containing reference marker position data d16, d16a, d16b, d16c Arranged reference shape data d17 Arranged individual shape data d18 Measurement accuracy d19 Arrangement accuracy d20 Object arrangement information d21 Determination result M1, M1a, M b, M1c, M1d, M1e Shape distortion conversion M2, M2a, M2b Shape placement shift conversion M3, M3a, M3b, M3c, M3d, M3e, M3f Shape placement matching transformation b1, b2, b3, b4, b5, b6 Three-dimensional position c1, c2, c3, c4 marker position shape v1, v2, v3, v4 marker position data f1, f2, f3, f4, f5, f6 marker c5 reference marker position shape c6 original reference marker position shape c7, c8 distortion-containing reference marker Position shape c6 ′, c7 ′, c8 ′ Position adjusted marker position shape v6 ′, v7 ′, v8 ′ Position adjusted marker position data

Claims (14)

Of the three-dimensional shape of the target object, the first shape data representing the three-dimensional shape of the first area and the second area for the first area and the second area having portions overlapping each other on the target object, and the second area Acquisition means for acquiring second shape data representing the three-dimensional shape of
Derived data generation means for performing a derived data generation process for generating a plurality of third shape data expressing a plurality of different geometric states based on the second shape data;
For each of the plurality of third shape data, a shape placement adjustment unit that generates a plurality of fourth shape data as a result of the processing by attempting a placement process on the first shape data in a predetermined coordinate system; ,
When trying to arrange the first shape data and the second shape data in a single coordinate system, a range including the first region and the second region of the object Determining means for determining whether or not the three-dimensional shape of the object can be reproduced with an accuracy within an allowable range, based on the mutual relationship of the shape arrangement of the plurality of fourth shape data;
A three-dimensional shape data processing apparatus comprising: Of the three-dimensional shape of the target object, the first shape data representing the three-dimensional shape of the first area and the second area for the first area and the second area having portions overlapping each other on the target object, and the second area Acquisition means for acquiring second shape data representing the three-dimensional shape of
Derived data generation means for performing a derived data generation process for generating a plurality of third shape data expressing a plurality of different geometric states based on the second shape data;
For each of the plurality of third shape data, a shape placement adjustment unit that tries to place the first shape data in a predetermined coordinate system;
When trying to arrange the first shape data and the second shape data in a single coordinate system, a range including the first region and the second region of the object A plurality of shapes for the second shape data when the plurality of third shape data is generated in the derived data generation process whether or not the three-dimensional shape of the object can be reproduced with an accuracy within an allowable range. Judgment that the variation information of the arrangement and the variation information of the plurality of shape arrangements for the plurality of third shape data in the arrangement adjustment process are determined based on the mutual relationship between the synthesized pieces of variation information of the plurality of shape arrangements Means,
A three-dimensional shape data processing apparatus comprising: The three-dimensional shape data processing apparatus according to claim 1 or 2,
The plurality of geometric states are:
(A) a displacement state of a shape arrangement corresponding to a three-dimensional position and posture in the predetermined coordinate system;
(B) a three-dimensional strain state;
A three-dimensional shape data processing apparatus characterized in that at least one of them is in a different state. A three-dimensional shape data processing apparatus according to any one of claims 1 to 3,
One of the plurality of third shape data is the second shape data itself, and the three-dimensional shape data processing device. A three-dimensional shape data processing apparatus according to any one of claims 1 to 3,
All of the plurality of third shape data represents a geometric state different from that of the second shape data. A three-dimensional shape data processing device according to any one of claims 1 to 5,
Shape data generating means for generating the first shape data and the second shape data used in the three-dimensional shape data processing device;
A three-dimensional shape data processing system comprising: The three-dimensional shape data processing system according to claim 6,
The shape data generating means generates at least one of the first shape data and the second shape data based on known shape data representing the shape of the object in terms of design numerical values. Original shape data processing system. The three-dimensional shape data processing system according to claim 6,
The shape data generation means is a shape data for the object measured by using a measuring device that measures at least one of the first shape data and the second shape data by measuring a three-dimensional shape. A three-dimensional shape data processing system generated based on A three-dimensional shape data processing system according to any one of claims 6 to 8,
When the determination means gives a determination result that denies the possibility of reproduction,
(A) The shape data generation means generates fifth shape data representing a three-dimensional shape of a third region on the object different from both the first region and the second region, and uses the shape of the object. Generate based on known shape data expressed numerically design,
(B) The three-dimensional shape data processing system, wherein the acquisition unit acquires the fifth shape data instead of the second shape data. A three-dimensional shape measurement system that generates shape data representing an overall three-dimensional shape of the object based on a plurality of individual shape data obtained by measuring the three-dimensional shape of the object from a plurality of directions,
A measuring device that performs three-dimensional measurement to obtain the plurality of individual shape data;
Arranging means for relatively arranging the measuring device and the object;
A three-dimensional shape data processing system according to any one of claims 6 to 9,
With
When the measuring apparatus acquires the plurality of individual shape data, the first region and the object of the object are obtained on the condition that the determination unit obtains a determination that affirms the possibility of reproduction. For the range including the second region, the measurement device and the object based on the arrangement relationship between the measurement device and the object corresponding to the first shape data and the second shape data, respectively. A three-dimensional shape measuring system characterized by relatively arranging the two. Among the three-dimensional shape of the object provided with a plurality of identification points, the first area and the second area having three or more identification points in a portion overlapping each other on the object, the first area in the first area First position data expressing the positions of three or more target identification points among the three or more identification points, and second position data expressing the positions of the three or more target identification points in the second area. And an acquisition means for acquiring
Derived data generating means for performing a derived data generating process for generating a plurality of third position data respectively representing a plurality of states having different distortion states of the distribution of the three or more target identification points based on the second position data When,
For each of the plurality of third position data, a shape arrangement adjusting unit that generates a plurality of fourth position data as a result of the process by trying an alignment process on the first position data in a predetermined coordinate system; ,
When trying to align the shape data representing the three-dimensional shape of the object with respect to the first area and the second area in a single coordinate system, the first area of the object Whether or not the three-dimensional shape of the object can be reproduced with accuracy within a predetermined allowable range for the range including the second region is based on the correlation of the shape arrangement with respect to the plurality of fourth position data. Determination means for determining
A three-dimensional shape data processing apparatus comprising: Among the three-dimensional shape of the object provided with a plurality of identification points, the first area and the second area having three or more identification points in a portion overlapping each other on the object, the first area in the first area First position data expressing the positions of three or more target identification points among the three or more identification points, and second position data expressing the positions of the three or more target identification points in the second area. And an acquisition means for acquiring
Derived data generating means for performing a derived data generating process for generating a plurality of third position data respectively representing a plurality of states having different distortion states of the distribution of the three or more target identification points based on the second position data When,
For each of the plurality of third position data, a shape arrangement adjusting unit that tries to arrange the first position data in a predetermined coordinate system;
When trying to align the shape data representing the three-dimensional shape of the object with respect to the first area and the second area in a single coordinate system, the first area of the object For each of the plurality of third position data in the alignment process, whether or not the three-dimensional shape of the object can be reproduced with accuracy within a predetermined allowable range for the range including the second region. Determination means for determining based on the mutual relationship of the variation information of the shape arrangement;
A three-dimensional shape data processing apparatus comprising: The three-dimensional shape data processing device according to claim 11 or 12,
First shape data representing the three-dimensional shape of the object for the first region, used for generating the first position data and the second position data used in the three-dimensional shape data processing apparatus; Shape data generating means for generating second shape data representing a three-dimensional shape of the object for the second region;
A three-dimensional shape data processing system comprising: A tertiary that generates shape data representing the overall three-dimensional shape of the object based on a plurality of individual shape data obtained by measuring the three-dimensional shape of the object provided with a plurality of identification points from a plurality of directions. An original shape measuring system,
A measuring device that performs three-dimensional measurement to obtain the plurality of individual shape data;
Arranging means for relatively arranging the measuring device and the object;
A three-dimensional shape data processing system according to claim 13;
With
When the measuring apparatus acquires the plurality of individual shape data, the first region and the object of the object are obtained on the condition that the determination unit obtains a determination that affirms the possibility of reproduction. For the range including the second region, the measurement device and the object based on the arrangement relationship between the measurement device and the object corresponding to the first shape data and the second shape data, respectively. A three-dimensional shape measuring system characterized by relatively arranging the two.

Images