JP5273091B2 - 3D shape measuring apparatus and 3D shape measuring method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To prevent deterioration in efficiency even if a position of a three-dimensional camera is varied, when measuring a three-dimensional shape while rotating a stage around a rotation axis. <P>SOLUTION: A controller 30 preliminarily memorizes a coordinate or a vector component of a fixed-point of a stage 10 obtained by measurement with a three-dimensional camera 20 arranged at a first camera point. The controller 30 obtains a camera coordinate conversion factor for coordinate-converting three-dimensional data expressing a three-dimensional shape measured by the three-dimensional camera 20 at a second camera position to three-dimensional data expressing a three-dimensional shape measured by the three-dimensional camera 20 at the first camera position, by using a coordinate or a vector component of the fixed-point of the stage 10 obtained by measurement with the three-dimensional camera 20 arranged at the second camera position different from the first camera position, and the preliminarily memorized coordinate or the vector component of the fixed-point of the stage 10. <P>COPYRIGHT: (C)2012,JPO&amp;INPIT

Description

  The present invention provides a three-dimensional shape measuring apparatus capable of acquiring a three-dimensional shape in all directions of a measuring object by placing the measuring object on a rotating stage and measuring the three-dimensional shape while rotating the stage. The present invention relates to a dimension shape measuring method.

  There are various methods for measuring the three-dimensional shape by irradiating the measurement object with light from the three-dimensional camera, but the measurable part is only the part where the measurement object is irradiated with light. In order to measure the three-dimensional shape in all directions of the measurement object, it is necessary to move the three-dimensional camera or the measurement object to measure the measurement object from various directions, and there are various methods. is there. As one of the methods, there is a method of placing a measurement object on a rotating stage and measuring a three-dimensional shape while rotating the stage as disclosed in Patent Document 1 below, for example.

  In this method, since the three-dimensional camera is rotated when viewed from the stage, the coordinate system is different for each preset stage rotation angle, and three-dimensional data for each rotation angle is set. A coordinate conversion coefficient for conversion into the three-dimensional data of the coordinate system (reference coordinate system) at the rotation angle obtained is acquired and stored for each rotation angle, and the three-dimensional data acquired at each rotation angle is converted to this coordinate. All three-dimensional data is converted into data in one coordinate system by conversion with coefficients. And in the method shown in the following Patent Document 1, a reference object having a fixed point is placed on the stage, the three-dimensional shape of the reference object is measured for each rotation angle, and three or more fixed point coordinates are acquired. The coordinate conversion coefficient is calculated. In the method disclosed in Patent Document 1, the rotational axis is only one of the normal directions of the plane of the stage, and the measurement object having a complicated structure at the upper and lower portions has a three-dimensional shape in all directions. Measurement may not be possible.

  On the other hand, in the three-dimensional shape measurement shown in Patent Document 2 below, a second rotary stage having a rotation axis different from that of the first rotary stage is placed on the first rotary stage, and the second rotary stage is placed on the second rotary stage. The measurement object is placed and the three-dimensional shape of the measurement object is measured while rotating the first and second rotating stages, that is, rotating the measurement object around two different rotation axes. Thereby, it is possible to measure a three-dimensional shape in all directions even with a measurement object having a complicated structure. Also in the method disclosed in Patent Document 2, coordinate conversion of three-dimensional data is necessary, and a coordinate conversion coefficient needs to be acquired for each rotation angle of each rotation axis. In the method disclosed in Patent Document 2, a coordinate conversion coefficient for each rotation angle of each rotation axis is obtained by detecting the position of each rotation axis in the reference coordinate system. Also in this method, as shown in Patent Document 1, a reference object having a fixed point is measured for a three-dimensional shape for each rotation angle of each rotation axis, and three or more fixed point coordinates are acquired. The coordinate conversion coefficient for each rotation angle of each rotation axis can also be calculated.

JP 2002-328014 A Special table 2000-509505 gazette

  The coordinate conversion coefficient for each rotation angle of the stage is constant if the position of the three-dimensional camera with respect to the stage does not change. However, when the position of the three-dimensional camera with respect to the stage is frequently changed, it is necessary to acquire a coordinate conversion coefficient for each rotation angle of the stage each time, and the measurement efficiency is deteriorated. As shown in Patent Document 2, when the stage is provided with two rotating shafts, the measurement efficiency is further deteriorated.

  The present invention has been made in order to cope with the above-described problem, and its purpose is to measure the three-dimensional shape of the measurement object while placing the measurement object on the stage and rotating the stage. An object of the present invention is to provide a three-dimensional shape measuring apparatus capable of efficiently measuring a three-dimensional shape of a measurement object even if the position of the camera is changed with respect to the stage. In addition, in the description of each constituent element of the present invention below, in order to facilitate understanding of the present invention, reference numerals of corresponding portions of the embodiment are described in parentheses, but each constituent element of the present invention is The present invention should not be construed as being limited to the configurations of the corresponding portions indicated by the reference numerals of the embodiments.

To achieve the above object, the present invention is characterized in that a stage (10) that can be rotated around two or more rotation axes and that can define a fixed point or vector for calculating a coordinate transformation coefficient, A three-dimensional camera (20) for measuring the three-dimensional shape of the object placed on the computer, and processing devices (30, 40) for processing three-dimensional data representing the three-dimensional shape measured by the three-dimensional camera. 3D representing the 3D shape measured by the 3D camera with the stage rotated from the reference position using the fixed point or vector of the stage acquired by the measurement of the 3D camera arranged at the first camera position. Two or more rotational coordinate conversion coefficients for converting data into three-dimensional data of a reference coordinate system representing a three-dimensional shape measured by a three-dimensional camera with the stage at the reference position . Rotation coordinate conversion coefficient acquisition means (coordinate conversion coefficient setting program shown in FIGS. 4A and 4B) that is acquired and stored for each of a plurality of rotation angles for each rotation axis, and the stage is rotated from the reference position Rotational coordinate conversion means for converting the three-dimensional data representing the three-dimensional shape of the measurement object measured by the three-dimensional camera into three-dimensional data in the reference coordinate system using the rotational coordinate conversion coefficient (steps S402 to S402 in FIG. 8). A three-dimensional shape measuring apparatus that combines the three-dimensional data of the measurement object measured from different viewpoints by rotating the stage with the three-dimensional data of the reference coordinate system, provided with S412, S418 to S422) Furthermore, the coordinates of the fixed point of the stage or the vector component obtained by the measurement of the three-dimensional camera arranged at the first camera position is stored. The point or vector storage means (step S106 in FIG. 4A), the coordinates of the fixed point of the stage or the vector component obtained by the measurement of the three-dimensional camera arranged at the second camera position different from the first camera position, The three-dimensional data representing the three-dimensional shape measured by the three-dimensional camera at the second camera position using the coordinates of the fixed point of the stage or the vector component stored by the vector storage means is obtained at the first camera position. Camera coordinate conversion coefficient acquisition means (steps S310 to S314 in FIG. 7) for acquiring a camera coordinate conversion coefficient for converting the coordinates into three-dimensional data representing the three-dimensional shape measured by the three-dimensional camera, and the second camera position. The three-dimensional data representing the three-dimensional shape of the measurement object measured by the three-dimensional camera is converted into the first camera position using the camera coordinate conversion coefficient. Camera coordinate conversion means for converting the three-dimensional data representing a three-dimensional shape of the measurement object measured by the three-dimensional camera and a (step S416 in FIG. 8) in the provision.

  In this case, the fixed point / vector storage means may store the coordinates or vector components of the fixed point of the stage in conjunction with the acquisition of the rotation coordinate conversion coefficient by the rotation coordinate conversion coefficient acquisition means.

In the present invention configured as described above, the camera coordinate conversion unit converts the three-dimensional data representing the three-dimensional shape of the measurement object measured by the three-dimensional camera at the second camera position by the camera coordinate conversion coefficient acquisition unit. Using the acquired camera coordinate conversion coefficient, it is converted into three-dimensional data representing the three-dimensional shape of the measurement object measured by the three-dimensional camera at the first camera position. Then, if the three-dimensional data converted by the camera coordinate conversion means is converted into the three-dimensional data of the reference coordinate system by the rotation coordinate conversion means using the rotation coordinate conversion coefficient, the stage is rotated at the second camera position. The three-dimensional data of the measurement object can be converted into the three-dimensional data of the reference coordinate system. Therefore, even if the position of the three-dimensional camera with respect to the stage is changed, the coordinate conversion coefficient for each rotation angle is obtained by performing only one measurement to obtain the coordinates of the fixed point or vector component defined on the stage. 3D shape measurement is possible in the same manner as above, and the measurement efficiency is kept good. Even when the stage is rotatable around two or more rotation axes, the measurement for obtaining the fixed point coordinates or vector components defined on the stage may be performed once. This is more efficient because there is no need to acquire a coordinate conversion coefficient for each rotation angle.

  Another feature of the present invention is that a fixed point or vector defined on the stage can be obtained by mounting an object capable of defining a fixed point or a vector in a fixed direction on the stage at a fixed position with respect to the stage. It is in. According to this, for example, an object that can define a fixed point or vector on the stage only when a mounting hole or screw hole is created on the stage and measurement is performed to acquire the coordinates of the fixed point or vector component defined on the stage. So that it does not get in the way when placing the measurement object on the stage.

  Another feature of the present invention is that the fixed point or vector defined in the stage is a vertex or plane normal vector in the shape of the stage. According to this, when acquiring the fixed point coordinates or vector components defined in the stage, it is only necessary to measure the stage as it is, so that the efficiency is further improved.

  Another feature of the present invention is that the fixed point or vector defined on the stage is a point or a straight line portion determined from a pattern drawn on the stage. Also by this, when acquiring the fixed point coordinates or vector components defined in the stage, it is only necessary to measure the stage as it is, which is efficient.

  Furthermore, in carrying out the present invention, the present invention is not limited to the invention of the three-dimensional shape measuring apparatus, and can also be carried out as an invention of a three-dimensional shape measuring method.

1 is an overall schematic diagram of a three-dimensional shape measurement system to which the present invention is applied. It is the figure which looked at the mode when setting a true sphere on a stage from the section direction of a stage. It is a figure which shows a mode when acquiring the coordinate transformation coefficient for every rotation angle of each rotating shaft. It is a flowchart which shows the first half part of the coordinate transformation coefficient setting program performed by a controller. It is a flowchart which shows the second half part of the coordinate transformation coefficient setting program performed by a controller. It is a flowchart which shows the detail of the reference | standard fixed point coordinate acquisition routine of FIG. 4A. It is a flowchart which shows the detail of the coordinate transformation coefficient acquisition routine of FIG. 4A, FIG. 4B, and FIG. It is a flowchart which shows the target object measurement program performed by a controller. It is a flowchart which shows the image data generation program run by a data processor. It is a figure which shows the relationship of a coordinate system when the position of the three-dimensional camera with respect to a stage is changed. It is a figure which shows the method when defining a fixed point and a vector in a stage. It is a figure which shows a mode when the position of each rotating shaft is detected.

a. Hardware Configuration Hereinafter, an embodiment of the present invention will be described with reference to the drawings. FIG. 1 is an overall schematic diagram of a three-dimensional shape measurement system according to an embodiment of the present invention. The three-dimensional shape measurement system includes a three-dimensional camera 20 that measures the three-dimensional shape of the measurement object OB placed on the stage 10.

  The stage 10 is formed in a disk shape and has a flat upper surface on which the measurement object OB is placed. Three holes 10a, 10b, and 10c for setting a true sphere are provided at three locations near the outer periphery of the upper surface of the stage 10. When measuring the measurement object OB, nothing is set in these holes 10a, 10b, 10c, but in the measurement for obtaining a coordinate conversion coefficient described later, these holes 10a, 10b, 10c. Is set with true spheres 11a, 11b, and 11c having different diameters. A cylindrical rod is attached to the true spheres 11a, 11b, and 11c, and the true spheres 11a, 11b, and 11c are set on the upper surface of the stage 10 by inserting the cylindrical rods into the holes 10a, 10b, and 10c. Is done. FIG. 2 is a view of the state in which the true sphere 11a (11b, 11c) with the rod is inserted into the hole 10a (10b, 10c) as seen from the cross-sectional direction. The diameters of the holes 10a, 10b, and 10c are different from each other by a small amount, and the diameters of the rods attached to the true spheres 11a, 11b, and 11c are also different from each other so as to fit the holes 10a, 10b, and 10c. Therefore, the true spheres 11a, 11b, and 11c having the fixed diameter are set at the fixed position of the stage 10. The rods provided in the holes 10a, 10b, 10c and the true spheres 11a, 11b, 11c are threaded, and the rods provided in the true spheres 11a, 11b, 11c are screwed into the holes 10a, 10b, 10c. May be set.

  The stage 10 is supported by a stage support device including plates 12, 13, 14, and 15, and motors 16, 17, and 18 are assembled to the stage support device. The plate 12 supports the plate 13 so as to be rotatable around the rotation axis A, and the motor 16 integrally rotates the plates 13, 14, 15 and the stage 10 around the rotation axis A by the rotation. Rotate positively or negatively within a predetermined range from 0 degrees. The plate 14 supports the plate 15 so as to be rotatable around the rotation axis B, and the motor 17 integrally rotates the plate 15 and the stage 10 around the rotation axis B from the reference position (rotation angle 0 degree). Rotate positive and negative by a predetermined range. The plate 15 fixedly supports the motor 18, and the motor 18 rotates the stage 10 around the rotation axis C by a free angle positive and negative from the reference position (rotation angle 0). A rotation angle sensor is built in each of the motors 16, 17, and 18, and the rotation angle sensor outputs a rotation angle around each rotation axis A, B, and C of the stage 10. The rotation axes A, B, and C are orthogonal to each other.

  The three-dimensional camera 20 is fixed to a movable support device 21 so as to be displaceable and rotatable. The three-dimensional camera 20 measures the three-dimensional surface shapes of the measurement object OB and the true spheres 11a, 11b, and 11c, and outputs three-dimensional data representing the measurement results. The three-dimensional camera 20 is not limited as long as it measures the three-dimensional surface shapes of the measurement object OB and the true spheres 11a, 11b, and 11c and outputs three-dimensional data representing the measured three-dimensional surface shapes. A 3D camera can also be used. In the present embodiment, a brief description will be given of measuring a three-dimensional surface shape of an object according to a triangulation method using laser light.

  In this three-dimensional camera 20, a virtual plane that is substantially perpendicular to the traveling direction of laser light emitted from a laser light source toward an object is assumed, and an X-axis direction and a Y-axis direction that are orthogonal to each other on the virtual plane. A large number of minute areas divided along the line are assumed. Then, the three-dimensional camera 20 sequentially irradiates the plurality of minute areas with laser light, sequentially detects the distance to the object surface defined by the minute areas by scattered light from the object as a Z-axis direction distance, Information on X, Y, and Z coordinates representing each divided area position obtained by dividing the surface of each object into minute areas is obtained, and the shape of the object surface facing the three-dimensional camera 20 is measured.

  Therefore, this three-dimensional camera has an X-axis direction scanner that changes the direction of outgoing laser light in the X-axis direction, a Y-axis direction scanner that changes the direction of outgoing laser light in the Y-axis direction, and an object surface. A distance detector that receives the reflected laser beam and detects the distance to the object surface. The X-axis direction scanner and the Y-axis direction scanner may be any mechanism that can change the optical path of the laser beam emitted from the laser light source independently in the X-axis direction and the Y-axis direction. Rotating around the axis in the direction and the Y-axis direction by an electric motor, or rotating the galvanometer mirror provided in the optical path of the emitted laser light and changing the direction around the axes in the X-axis direction and the Y-axis direction by the electric motor Can be used. As the distance detector, an imaging lens that rotates following the optical path of the emitted laser light and collects the reflected laser light (scattered light) reflected (scattered) on the object surface and the condensed laser light It is possible to use a mechanism that includes a line sensor in which a plurality of light receiving elements such as CCDs that receive light are arranged in a line, and that detects the distance to the object surface by the light receiving position of the reflected laser light by the line sensor.

  Therefore, such a three-dimensional camera 20 uses information about X, Y, and Z coordinates representing the divided area positions obtained by dividing the surface of the object into minute areas with respect to the reference direction of the laser beam emitted by the X-axis direction scanner. The inclination θx in the X-axis direction, the inclination θy in the Y-axis direction with respect to the reference direction of the laser beam emitted by the Y-axis direction scanner, and the distance Lz to the object surface by the distance detector are Detection is performed for each of a large number of minute areas divided along the Y-axis direction. More specifically, the inclinations θx and θy in the X-axis and Y-axis directions are rotation angles from the reference position of the electric motor. Further, the distance Lz to the object surface is the light receiving position of the reflected laser beam in the line sensor. The inclinations θx, θy and the distance Lz are converted into X, Y, Z coordinate values and output. The conversion to the X, Y, Z coordinate values is performed by the controller 30 and the data processing device 40 that receive the output from the three-dimensional camera 20, or the three-dimensional camera 20, the controller 30, and the data processing device 40. Alternatively, a converter may be provided between the two.

  A controller 30 and a data processing device 40 are connected to the three-dimensional camera 20. The controller 30 is constituted by a computer device having a large-capacity nonvolatile memory such as a CPU, ROM, RAM, and hard disk, and in accordance with an instruction from an input device 31 including a keyboard, the three-dimensional camera 20 and the motors 16, 17, 18. And the operation of the data processing device 40 is controlled. Specifically, the controller 30 commands the three-dimensional camera 20 to perform measurement by inputting from an input device by an operator or executing a built-in program, and designates motors 16, 17, and 18 that rotate the stage 10. Just rotate. In addition, rotation angles about the rotation axes A, B, and C of the stage 10 are input from the built-in rotation angle sensors from the motors 16, 17, and 18, and rotation angle data is output to the data processing device 40. Furthermore, the controller 30 calculates a coordinate conversion coefficient by program processing.

  The data processing device 40 is also composed of a computer device having a large-capacity non-volatile memory such as a CPU, ROM, RAM, and hard disk, and performs coordinate conversion of the three-dimensional data from the three-dimensional camera 20 according to a command from the controller 30. The three-dimensional image data is created by executing the process including the three-dimensional image data, and the three-dimensional image is displayed on the display device 41 using the three-dimensional image data. Further, the data processing device 40 stores data necessary for calculation of the three-dimensional shape of the measurement object OB, such as coordinate conversion coefficients and fixed point coordinates, in a RAM, a nonvolatile memory, or the like.

b. Setting of Coordinate Conversion Coefficient In the three-dimensional shape measurement system configured as described above, the operator acquires the coordinate conversion coefficient for each rotation angle around each rotation axis A, B, C only for the first time, and acquires the acquired coordinate conversion coefficient. To supply the data to the data processing device 40.

  First, as shown in FIG. 3, the operator assembles the rod-attached true spheres 11 a, 11 b, and 11 c in the three holes 10 a, 10 b, and 10 c near the outer periphery of the stage 10. The spheres 11a, 11b, and 11c have different radii. The reason why the spheres 11a, 11b, and 11c have different radii is that the three-dimensional data of the specified spheres 11a, 11b, and 11c is extracted from the three-dimensional data acquired by the three-dimensional shape measurement by the three-dimensional camera 20. This is to make it easier. The worker also supports the three-dimensional camera 20 on the support device 21 and causes the three-dimensional camera 20 to measure the three-dimensional shapes of the true spheres 11a, 11b, and 11c.

  Next, the operator operates the input device 31 to cause the controller 30 to execute the coordinate conversion coefficient setting program shown in FIGS. 4A and 4B. The execution of the coordinate conversion coefficient setting program is started in step S100 of FIG. 4A, and the controller 30 determines that the rotation angle of the stage 10 around the rotation axes A, B, C is the reference rotation angle (that is, 0 degrees) in step S102. ), The motors 16, 17, and 18 are controlled to rotate. Hereinafter, this state is referred to as a reference position of the stage 10, and the three-dimensional data obtained by the three-dimensional camera 20 at the position shown in FIGS. It is called data. Next, in step S104, the controller 30 executes a process for inputting the feature of the detection target object. In this input process, the display device 31 displays the feature of the detection target object so as to prompt the operator to input the feature of the detection target object. In the case of this embodiment, the true spheres 11a, 11b, and 11c are spheres, and the input of the radius of the true spheres 11a, 11b, and 11c is prompted. When the operator operates the input device 31 to input the feature, the controller 30 temporarily stores the feature and executes a reference fixed point coordinate acquisition routine in step S106. This reference fixed point coordinate acquisition routine is a process of acquiring the coordinates of three fixed points set on the stage 10 at the reference position, and in this embodiment, the center points of the true spheres 11a, 11b, and 11c assembled to the stage 10. This is a process for acquiring the X, Y, and Z coordinate values. These X, Y, and Z coordinate values are coordinate values based on the reference coordinate system.

  The details of this reference fixed point coordinate acquisition routine are shown in FIG. 5, and the controller 30 starts the execution of this reference fixed point coordinate acquisition routine in step S200. After starting the execution of the reference fixed point coordinate acquisition routine, the controller 30 instructs the three-dimensional camera 20 to measure the three-dimensional shape including the true spheres 11a, 11b, and 11c on the stage 10 in step S202. Three-dimensional data as measurement results are inputted from the three-dimensional camera 20 and temporarily stored. Next, in step S204, the controller 30 represents the three-dimensional shape of the surface of each of the true spheres 11a, 11b, and 11c for each of the true spheres 11a, 11b, and 11c from all the input three-dimensional data. Each three-dimensional data is extracted. This type of three-dimensional data extraction technique is a conventionally known technique (see, for example, Japanese Patent Application Laid-Open Nos. 2004-333371 and 2005-249402), but will be briefly described.

<Extraction of 3D data>
Using the data representing the characteristics (sphere and radius) of the input detection target object, unit block and search block size setting processing is executed. The unit block is the minimum block for moving the search block in order to specify the position where the detection target object exists, and is formed in a cube, for example. The size of the unit block is set to be small to some extent so that it can be confirmed that a part of the detection target object exists. The search block is used to specify a position where the detection target object is included, and is formed in a cube in this embodiment. The size of this search block can be set to be as small as possible while including all of the detection target objects. However, being able to include this detection target object means including all of the unit blocks including even a part of the detection target object.

  Next, search area blocking processing is executed. This search area blocking process is a process of dividing an area that may contain the detection target object in the measurement target area into unit blocks. Basically, a space where three-dimensional data exists in the measurement target space is three-dimensionally divided into unit blocks. The division is performed by a method of arranging unit blocks along the coordinate axes of the X, Y, and Z coordinates. Then, it is checked whether or not there is a predetermined number of three-dimensional data for each of the divided unit blocks, and unit blocks having a predetermined number of three-dimensional data are extracted. Next, detection processing of a search block position that may include the detection target object is performed. In detection of the search block position, the set search block is sequentially moved in the X-axis, Y-axis, and Z-axis directions with the unit block as a unit in the region divided into the unit blocks. For each movement, the number of unit blocks extracted and included in the search block after movement is calculated. If the number of unit blocks is within a predetermined range, the same position is detected as the corresponding search block position. Since the detection target object is a sphere, a position where the 3D data is obtained by being located on the side facing the 3D camera 20 and a 3D data being obtained on the side not facing the 3D camera 20 are obtained. Since the ratio of the non-existing portions is almost the same, if the radius of the sphere is specified, the number of unit blocks included in the search block is set. Thereby, a sphere having a radius smaller than that of the detection target object can be excluded as the number of unit blocks is outside the predetermined range.

  Next, it is determined whether or not the three-dimensional data included in the search block at the detected position matches a sphere that is the shape of the detection target object. Extract. Specifically, all the three-dimensional data (X, Y, Z coordinate values) in the corresponding search block are respectively substituted into x, y, z on the left side of the following equation 1 representing the sphere, and the minimum The unknowns a, b, c, and d are calculated using the square method. In this case, a, b, and c represent X, Y, and Z coordinate values of the sphere center represented by the three-dimensional data, respectively, and d represents the radius of the sphere. Next, for each three-dimensional data (X, Y, Z coordinate value) in the corresponding search block, the same three-dimensional data (X, Y, Z coordinate value) and the calculated values a, b, c are obtained. Substituting into Equation 1 below, the value d (distance from the center of the sphere) is calculated for each three-dimensional data. The difference between the calculated value d (the radius of the sphere) and the radius input as the feature is within a predetermined determination value, and the deviation of the distance from the sphere center of each three-dimensional data is within the predetermined determination value. If so, the three-dimensional data in the search block is determined to match, and the three-dimensional data is extracted. On the other hand, if the difference or deviation is not within a predetermined discriminant value, the 3D data in the search block is not matched and the 3D data is not extracted. Thereby, the three-dimensional data regarding the true spheres 11a, 11b, and 11c are extracted for each of the true spheres 11a, 11b, and 11c.

(Equation 1)
(x−a) ² + (y−b) ² + (z−c) ² −d ² = 0

  Here, the description returns to the reference fixed point coordinate acquisition routine of FIG. After the process of step S204, the controller 30 determines in step S206 the central coordinates of the true spheres 11a, 11b, 11c for each of the true spheres 11a, 11b, 11c, that is, X, Y of the three fixed points provided on the stage 10. , Z coordinate value is calculated. Specifically, for each of the true spheres 11a, 11b, and 11c, three-dimensional data (X, Y, and Z coordinate values) related to the true spheres 11a, 11b, and 11c extracted by the process of step S204 are expressions representing the sphere. And the unknowns a, b, and c are respectively calculated using the least-squares method by substituting for the values of x, y, and z of the following Equation 1 and using the radius input in step S104 as the value d. The unknowns a, b, c calculated for each of the true spheres 11a, 11b, 11c are the center coordinates of the true spheres 11a, 11b, 11c, that is, the X, Y, Z coordinate values (Xa0, Ya0, Za0), (Xb0, Yb0, Zb0), (Xc0, Yc0, Zc0).

  After the process of step S206, in step S208, the controller 30 stores the X, Y, Z coordinate values (Xa0, Ya0, Za0), (Xb0, Yb0, Zb0) of the three fixed points in a non-volatile memory such as a hard disk. ), (Xc0, Yc0, Zc0) are stored as coordinates of three fixed points when the stage 10 is at the reference position, that is, reference fixed point coordinates (coordinates of fixed points in the reference coordinate system). And the controller 30 complete | finishes execution of a reference | standard fixed point coordinate acquisition routine in step S210, and returns to step S108 of FIG. 4A.

  In step S108, the controller 30 initializes the variable n to “0”, and executes the circulation process including steps S110 to S116. The variable n is a variable for sequentially changing the rotation angle of the stage 10 by a predetermined angle Q (for example, 1 degree) for each circulation process. In the circulation process including steps S110 to S116, the controller 30 adds “1” to the variable n in step S110, and the rotation angle n · Q around the rotation axis A of the stage 10 is determined in advance in step S112. It is determined whether it is larger than the upper limit angle Amax. If the rotation angle n · Q is not larger than the upper limit angle Amax, the controller 30 determines “No” in step S112, and controls the motor 16 in step S114 to move the stage 10 around the rotation axis A. The rotation angle around the rotation axis A of the stage 10 is set to n · Q degrees. Then, the controller 30 executes a coordinate conversion coefficient acquisition routine in step S116.

  This coordinate conversion coefficient acquisition routine is shown in detail in FIG. 6, and the controller 30 starts executing the coordinate conversion coefficient acquisition routine in step S250. After starting the execution of this coordinate conversion coefficient acquisition routine, the controller 30 executes the processes of steps S252 to S256 similar to the above-described steps S202 to 206 of FIG. 5 to obtain the center coordinates of the true spheres 11a, 11b, and 11c, that is, 3 The X, Y, Z coordinate values (Xa1, Ya1, Za1), (Xb1, Yb1, Zb1), (Xc1, Yc1, Zc1) of the fixed points are calculated. However, in this case, the stage 10 is at a position rotated by n · Q degrees from the reference position about the rotation axis A, and the calculated X, Y, Z coordinate values (Xa1, Ya1, Za1) of the three fixed points are calculated. (Xb1, Yb1, Zb1) and (Xc1, Yc1, Zc1) are the three fixed points X, Y, and X calculated using the three-dimensional data measured by the three-dimensional camera 20 with the stage 10 in the reference position. It is different from the Z coordinate values (Xa0, Ya0, Za0), (Xb0, Yb0, Zb0), (Xc0, Yc0, Zc0).

  After the processing of steps S252 to S256, the controller 30 determines the coordinates of the three-dimensional data measured by the three-dimensional camera 20 in a state where the stage 10 is rotated about the rotation axis A by n · Q degrees from the reference position in step S258. A coordinate conversion coefficient for converting the value into the coordinate value (coordinate value of the reference coordinate system) of the three-dimensional data measured by the three-dimensional camera 20 in a state where the stage 10 is at the reference position is calculated. This coordinate conversion coefficient is calculated by calculating the X, Y, Z coordinate values (Xa1, Ya1, Za1) of three fixed points in a state where the calculated stage 10 is rotated around the rotation axis A by n · Q degrees from the reference position. , (Xb1, Yb1, Zb1), (Xc1, Yc1, Zc1) are the X, Y, Z coordinate values (Xa0, Ya0, Za0), (Xb0, Yb0) of the three fixed points when the stage 10 is at the reference position. , Zb0), (Xc0, Yc0, Zc0) is calculated, and both X, Y, Z coordinate values (Xa1, Ya1, Za1), (Xb1, Yb1, Zb1), (Xc1, Yc1, Zc1), (Xa0, Ya0, Za0), (Xb0, Yb0, Zb0), (Xc0, Yc0, Zc0). This type of coordinate conversion coefficient calculation is a conventionally known technique (see, for example, Japanese Patent Application Laid-Open Nos. 2005-352630 and 2005-249402), but will be briefly described.

<Calculation of coordinate transformation coefficient>
First, the three-dimensional data obtained in a state where the stage 10 is rotated about the rotation axis A by n · Q degrees from the reference position is the three-dimensional data of the first coordinate system, and the reference coordinate system is the second coordinate system. Assume that Therefore, the coordinate axes of the first coordinate system are rotated by desired angles around the X, Y, and Z axes, respectively, and the origin of the first coordinate system is set to a in the X axis direction, the Y axis direction, and the Z axis direction, respectively. , B, c should be the same as the coordinate axes of the second coordinate system. Therefore, when the coordinate value of one point in the first coordinate system is (x ′, y ′, z ′) and the coordinate value of the same point in the second coordinate system is (x, y, z), the following formula 2 is established. In addition, the matrix M in the equation 2 is expressed by the following equation 3. The calculation of the coordinate conversion coefficient is performed by calculating the matrix values g ₁₁ , g ₁₂ , g ₁₃ , g ₂₁ , g ₂₂ , g ₂₃ , g ₃₁ , g ₃₂ , g ₃₃ and the matrix value a in the equations 2 and 3. , B, c.

  Here, X, Y, Z coordinate values (Xa1, Ya1, Za1), (Xb1, Yb1, Zb1), (Xc1, Yc1, Zc1) of the three fixed points in the first coordinate system and three of the second coordinate system. Substituting the X, Y, Z coordinate values (Xa0, Ya0, Za0), (Xb0, Yb0, Zb0), (Xc0, Yc0, Zc0) of the fixed points into the above equation 2, the following equations 4-6 are established. To do.

  When Equation 4 is transformed, the following simultaneous equations of Equation 7 are established.

  Further, the component in the first coordinate system of the normal vector of the plane including the three fixed points is (α ′, β ′, γ ′), and the normal vector of the same plane in the second coordinate system is (α, β, If γ), the following equation 8 is generally established if the two normal vectors have the same magnitude. The matrix M in the equation 8 is expressed by the equation 3.

  A normal vector of a plane including all fixed points in the second coordinate system and the first coordinate system, a vector directed from the center of the true sphere 11b (second fixed point) to the center of the true sphere 11a (first fixed point), and a true sphere Assuming that the vector is formed by the outer product with the vector from the center (fixed point) of 11c to the center of the true sphere 11b (second fixed point), the normal vector in the second coordinate system and the normal vector in the first coordinate system are as follows: 9 is represented.

  When the formula 9 is applied to the formula 8, the following formula 10 is established.

  If the first equation of the equation 10 is added to the equation 7, the simultaneous equations of the following equation 11 are obtained.

By solving the simultaneous equations of Equation 11, matrix values g ₁₁ , g ₁₂ , and g ₁₃ can be calculated. Also, with respect to the equations (5) and (6), if the equation is modified like the simultaneous equations of the equation (7) and the second or third equation of the equation (10) is added, the simultaneous equations similar to the simultaneous equations of the equation (11) are obtained. Since it is possible to solve this, matrix values g ₂₁ , g ₂₂ , g ₂₃ and matrix values g ₃₁ , g ₃₂ , g ₃₃ can be calculated. Then, by substituting these calculated matrix values into the equations 4 to 6, matrix values a, b, and c can be calculated. As a result, the coordinate values (x ′, y ′, z ′) of the first coordinate system, that is, the coordinate system of the three-dimensional data in a state where the stage 10 is rotated by n · Q degrees from the reference position around the rotation axis A are A coordinate conversion coefficient for conversion to a coordinate value (x, y, z) in the two coordinate system, that is, the reference coordinate system is calculated.

Here, the description returns to the coordinate conversion coefficient acquisition routine of FIG. After calculating the matrix values g ₁₁ , g ₁₂ , g ₁₃ , g ₂₁ , g ₂₂ , g ₂₃ , g ₃₁ , g ₃₂ , g ₃₃ and the matrix values a, b, c and the matrix values in step S258, the controller In step S260, 30 stores the calculated coordinate conversion coefficient in a non-volatile memory such as a hard disk, together with the rotation angle about the rotation axis A being n · Q degrees. And the controller 30 complete | finishes execution of a coordinate transformation coefficient acquisition routine in step S262, and returns to step S110 of FIG. 4A.

  The controller 30 adds “1” to the variable n in step S110, and unless the rotation angle n · Q around the rotation axis A of the stage 10 becomes larger than the predetermined upper limit angle Amax, “ It determines with "No" and performs the process of step S114, S116 mentioned above. As a result, while the stage 10 is rotated by Q degrees in the positive direction around the rotation axis A, the above-described coordinate conversion coefficient is calculated and stored in the nonvolatile memory. As a result, in the nonvolatile memory, the coordinate conversion coefficient in a state where the stage 10 is rotated in the positive direction by Q degrees around the rotation axis A from the reference position to the upper limit angle Amax is the rotation angle n · Q degrees of the stage 10. Each is stored in correspondence. In the present embodiment, since the Q degree is set to 1 degree, the coordinate conversion coefficient for each degree is stored in the nonvolatile memory. If this Q degree is set to a value other than 1 degree, a coordinate conversion coefficient for each predetermined Q degree other than 1 degree is stored in the nonvolatile memory. The same applies to other coordinate transformation coefficients described later. When the rotation angle n · Q of the stage 10 around the rotation axis A becomes larger than the upper limit angle Amax, the controller 30 determines “Yes” in step S112 and proceeds to step S118.

  In step S118, the controller 30 initializes the variable n to “0” and executes the circulation process including steps S120 to S126. The circulation process including steps S120 to S126 is the same as the process including steps S110 to S116 described above except that the processes of steps S122 and S124 are slightly different from the processes of steps S112 and S114 described above. In step S122, it is determined whether the rotation angle −n · Q around the rotation axis A of the stage 10 is smaller than a predetermined lower limit angle −Amax. In step S124, the motor 16 is controlled, the stage 10 is rotated in the negative direction around the rotation axis A, and the rotation angle around the rotation axis A of the stage 10 is set to -n · Q degrees. As a result, by the circulation process of steps S120 to S126, the rotation angle of the stage 10 around the rotation axis A extends from the reference position (0 degree) to the substantially lower limit angle −Amax, and the 3D by the three-dimensional camera 20 every Q degrees. A coordinate conversion coefficient for converting the dimension data into the three-dimensional data of the reference coordinate system is calculated and stored in the nonvolatile memory in correspondence with the rotation angle −n · Q around the rotation axis A of the stage 10. When the rotation angle −n · Q of the stage 10 around the rotation axis A becomes smaller than the lower limit angle −Amax, the controller 30 determines “Yes” in step S122 and proceeds to step S128.

  In step S128, the controller 30 controls the rotation of the motor 16 so that the rotation angle of the stage 10 around the rotation axis A becomes the reference rotation angle (that is, 0 degree). In this case, since the rotation angles of the stage 10 around the rotation axis B and the rotation axis C have already been set to the reference rotation angle (that is, 0 degrees) by the process of step S102 described above, the stage 10 is returned to the reference position. become. Next, the controller 30 initializes the variable n to “0” in step S <b> 130, and executes processing including steps S <b> 132 to S <b> 148. The process consisting of steps S132 to S148 is the same as the process consisting of steps S110 to S126 described above, except that the processes of steps S134, S136, S144, and S146 are slightly different from the processes of steps S112, S114, S122, and S124 described above. is there.

  In step S134, it is determined whether the rotation angle n · Q around the rotation axis B of the stage 10 is larger than a predetermined upper limit angle Bmax. In step S136, the motor 17 is controlled, the stage 10 is rotated in the positive direction around the rotation axis B, and the rotation angle around the rotation axis B of the stage 10 is set to n · Q degrees. In step S144, it is determined whether the rotation angle −n · Q around the rotation axis B of the stage 10 is smaller than a predetermined lower limit angle −Bmax. In step S146, the motor 16 is controlled, the stage 10 is rotated in the negative direction around the rotation axis B, and the rotation angle around the rotation axis B of the stage 10 is set to -n · Q degrees. As a result, three-dimensional data from the three-dimensional camera 20 is obtained at every Q degrees by the processing of steps S132 to S148 over the rotation angle of the stage 10 around the rotation axis B from the substantially lower limit angle −Bmax to the almost upper limit angle Bmax. A coordinate conversion coefficient for conversion into the three-dimensional data of the reference coordinate system is calculated and stored in the nonvolatile memory in correspondence with the rotation angles n · Q and −n · Q around the rotation axis B of the stage 10. When the rotation angle −n · Q of the stage 10 around the rotation axis B becomes smaller than the lower limit angle −Bmax, the controller 30 determines “Yes” in step S144 and proceeds to step S150 in FIG. 4B.

  In step S150, the controller 30 controls the rotation of the motor 17 so that the rotation angle of the stage 10 around the rotation axis B becomes the reference rotation angle (that is, 0 degree). In this case, since the rotation angles of the stage 10 around the rotation axis A and the rotation axis C have already been set to the reference rotation angle (that is, 0 degrees) by the processing in steps S128 and S102 described above, the stage 10 is returned to the reference position. Will be. Next, the controller 30 initializes the variable n to “0” in step S150, and executes the circulation process including steps S154 to S160. The circulation process consisting of steps S154 to S160 is the same as the above-described steps S110 to S116 (or steps S132 to S138) except that the processes of steps S156 and S158 are slightly different from the processes of steps S112 and S114 (or steps S134 and S136). ).

In step S156, it is determined whether the rotation angle n · Q around the rotation axis C of the stage 10 is greater than a predetermined angle 360 degrees. In step S158, the motor 18 is controlled, the stage 10 is rotated in the positive direction around the rotation axis C, and the rotation angle around the rotation axis C of the stage 10 is set to n · Q degrees. As a result, the rotation process around the rotation axis C of the stage 10 extends from the reference angle (0 degree) to almost 360 degrees by the circulation process including steps S154 to S160, and the three-dimensional camera 20 performs the three-dimensional operation every Q degrees. A coordinate conversion coefficient for converting the data into the three-dimensional data of the reference coordinate system is calculated and stored in the nonvolatile memory in correspondence with the rotation angle n · Q around the rotation axis C of the stage 10. When the rotation angle n · Q of the stage 10 around the rotation axis C becomes larger than the predetermined angle 360 degrees, the controller 30 determines “Yes” in step S156 and proceeds to step S162. In step S162, the coordinate transformation coefficients (matrix values g ₁₁ , g ₁₂ , g ₁₃₎ relating to the rotation axes A, B, C stored by the processing of the coordinate transformation coefficient acquisition routine in steps S116, S126, S138, S148, S160. , G ₂₁ , g ₂₂ , g ₂₃ , g ₃₁ , g ₃₂ , g ₃₃ and matrix values a, b, c) are output to the data processor 40. After the process of step S162, the controller 30 ends the execution of the coordinate conversion coefficient setting program in step S164.

The coordinate transformation coefficients related to the rotation axes A, B, and C output to the data processing device 40 are stored in a non-volatile memory such as a hard disk of the data processing device 40 by program processing (not shown). Here, coordinate conversion coefficients related to the rotation axes A, B, and C stored in the data processing device 40 will be organized. The coordinate conversion coefficient for the rotation axis A is determined by rotating the stage 10 around the rotation axis A by a minute predetermined angle Q (for example, 1 degree) from the lower limit angle −Amax to the upper limit angle Amax. When measured at 20, the three-dimensional data as a result of measurement by the three-dimensional camera 20 for each rotational position is in a state where the rotation angle around the rotation axis A of the stage 10 is at a reference angle (for example, 0 degrees) (stage 10 Matrix values g ₁₁ , g ₁₂ , g ₁₃ , g ₂₁ , g ₂₂ , g ₂₃ , g for conversion into the three-dimensional data of the object on the stage 10 measured by the three-dimensional camera 20 in a state where the image is at the reference position). _A matrix M composed of ₃₁ , g ₃₂ , and g ₃₃ and matrix values a, b, and c. Here, the rotation axis B, and to distinguish them from the coordinate transformation coefficient regarding rotation around C, since, relates the coordinate conversion factor for rotation around the rotation axis A, while the matrix M _A to the matrix M, the matrix value Let a, b, c be matrix values a _A , b _A , c _A.

Further, the coordinate conversion coefficient for the rotation axis B is 3 by rotating the stage 10 around the rotation axis B by a minute predetermined angle Q (for example, 1 degree) from the lower limit angle −Bmax to the upper limit angle Bmax. When measured with the three-dimensional camera 20, three-dimensional data, which is a measurement result of the three-dimensional camera 20 for each rotational position, is in a state where the rotation angle around the rotation axis B of the stage 10 is at a reference angle (for example, 0 degrees) ( Matrix values g ₁₁ , g ₁₂ , g ₁₃ , g ₂₁ , g ₂₂ , g ₂₃ for converting into three-dimensional data of the object on the stage 10 measured by the three-dimensional camera 20 in a state where the stage 10 is at the reference position). , G ₃₁ , g ₃₂ , and g _{33 are} matrix M and matrix values a, b, and c. Here, the rotation axis A, in order to distinguish it from the coordinate conversion factor for rotation around C, since, relates the coordinate conversion factor for rotation around the rotation axis B, while the matrix M and the matrix M _B, the matrix value Let a, b, c be matrix values a _B , b _B , c _B.

Further, the coordinate conversion coefficient regarding the rotation axis C is determined by rotating the stage 10 around the rotation axis C by a minute predetermined angle Q (for example, 1 degree) from 0 degrees to 360 degrees, and moving the object on the stage 10 to the three-dimensional camera 20. When the three-dimensional data as a result of measurement by the three-dimensional camera 20 for each rotational position is measured, the rotation angle around the rotation axis C of the stage 10 is at a reference angle (for example, 0 degrees) (the stage 10 is Matrix values g ₁₁ , g ₁₂ , g ₁₃ , g ₂₁ , g ₂₂ , g ₂₃ , g ₃₁ for conversion into three-dimensional data of the object on the stage 10 measured by the three-dimensional camera 20 in a state of being at the reference position). , G ₃₂ , g ₃₃ and matrix values a, b, c. Here, in order to distinguish it from the rotational axis A, the coordinate transformation coefficients for rotation around of B, and relates the coordinate conversion factor for rotation around the rotation axis C, while the matrix M and the matrix M _C, the matrix value Let a, b, c be matrix values a _C , b _C , c _C.

c. Measurement of Measurement Object Next, an operation for measuring the measurement object OB with the three-dimensional camera 20 will be described. In this case, the operator inputs to the controller 30 that he / she operates the input device 31 to measure the measurement object OB. The controller 30 starts executing the object measurement program in step S300 of FIG. Measuring the measurement object OB is supplied to the data processing device 40, and the data processing device 40 starts executing the image data generation program in step S400. First, as shown in FIG. 1, the measurement object OB is placed and fixed on the stage 10. First, the case where the three-dimensional camera 20 is fixed to the position at the time of executing the above-described coordinate transformation coefficient setting program and the three-dimensional shape of the measurement object OB is measured using the three-dimensional camera 20 will be described.

  When the operator operates the input device 31 to indicate the rotational positions of the stage 10 around the rotation axes A, B, and C, the controller 30 controls the motors 16, 17, and 18 by executing a program (not shown). Then, the stage 10 is rotated around the rotation axes A, B, and C to the rotation position instructed by the operator. Then, the operator operates the input device 31 to instruct the controller 30 to start measuring the measurement object OB. In response to the measurement start instruction, the controller 30 determines “Yes” in step S302 of FIG. 7 and notifies the data processing apparatus 40 that the measurement start instruction is issued in step S304. Next, in step S306, the controller 30 inputs rotation angles around the rotation axes A, B, and C of the stage 10 from rotation angle sensors built in the motors 16, 17, and 18, respectively. The data is output to the processing device 40. Next, in step S308, the controller 30 instructs the three-dimensional camera 20 to start measurement, and executes processes after step S310 described later. In response to this measurement start instruction, the three-dimensional camera 20 starts measuring the three-dimensional shape of the measurement object OB on the stage 10 and sequentially outputs three-dimensional data representing the measured three-dimensional shape.

  On the other hand, in response to the input of the object measurement start instruction in step S304 by the controller 30, the data processing device 40 determines “Yes” in step S402 in FIG. Waiting for the input of rotation angles around the rotation axes A, B, and C of the stage 10. When the rotation angle is input from the controller 30 in the process of step S306, the controller 30 inputs the rotation angle output from the controller 30 and stores it in the RAM or nonvolatile memory in step S406. Next, the data processing device 40 waits for the start of input of the three-dimensional data from the three-dimensional camera 20 in step S408. When the three-dimensional data starts to be supplied from the three-dimensional camera 20, the data processing device 40 sequentially inputs the three-dimensional data related to the measurement object OB sequentially output from the three-dimensional camera 20 in step S410. Or it memorize | stores in a non-volatile memory. The process of step S410 continues to be executed until the input of the three-dimensional data from the three-dimensional camera 20 is completed. When the input of the three-dimensional data from the three-dimensional camera 20 is completed, the data processing device 40 determines “Yes” in step S412, and proceeds to step S414 and the subsequent steps. At this stage, all the three-dimensional data acquired by the three-dimensional camera 20 is stored in the RAM or nonvolatile memory.

  In step S414, it is determined whether or not the camera movement flag CMF is “1”. When the camera movement flag CMF is “0”, the three-dimensional data of the measurement object OB is obtained while the three-dimensional camera 20 is fixed at the same position as when the coordinate conversion coefficient relating to the rotation about the rotation axes A, B, and C is acquired. This represents a state to be acquired, and represents a state in which the three-dimensional camera 20 is moved from the position by “1” to acquire the three-dimensional data of the measurement object OB, and is initially set to “0”. Now, since the three-dimensional camera 20 is fixed in a state where the coordinate conversion coefficient for the rotation about the rotation axes A, B, and C of the stage 10 is acquired, the camera movement flag is “0”, and the controller 30 It is determined as “No” in S414, and the processing after Step S418 is performed.

In step S418, the data processing device 40 uses the three-dimensional data input and stored from the three-dimensional camera 20 in the process of step S410, and the rotation angle around the rotation axis A of the stage 10 is a reference angle (for example, , 0 degrees) (in a state where the stage 10 is at the reference position), it is converted into the three-dimensional data of the object on the stage 10 acquired by the three-dimensional camera 20. In this case, the stage 10 is rotated around the rotation axis A from the lower limit angle −Amax to the upper limit angle Amax, acquired, and stored in the nonvolatile memory, the coordinate conversion coefficient (matrix M _A ) for each minute predetermined angle Q. Of the matrix value g and the matrix values a _A , b _A , c _A ), the coordinate conversion coefficient (of the matrix M _A ) corresponding to the rotation angle around the rotation axis A that is input from the controller 30 and temporarily stored. The matrix value g and the matrix values a _A , b _A , c _A ) are selected. The matrix value g and the matrix value g in the following description simply describe the matrix values g ₁₁ , g ₁₂ , g ₁₃ , g ₂₁ , g ₂₂ , g ₂₃ , g ₃₁ , g ₃₂ , and g ₃₃ described above. It is a thing. Then, the three-dimensional data temporarily stored is input from the three-dimensional camera 20, the selected coordinate transformation coefficients (matrix value g and the matrix values a _A of the matrix _{M A, b A, c A} ) using Coordinate conversion is performed by executing the following equation (12). Note that the coordinate values (X, Y, Z) in the following equation 12 are the three-dimensional data actually acquired by the three-dimensional camera 20, and the coordinate values (X _A , Y _A , Z _A ) This is the three-dimensional data of the object on the stage 10 measured by the three-dimensional camera 20 with the rotation angle around the rotation axis A being at the reference angle.

In this case, if the rotation mechanism of the stage 10 is a mechanism in which the rotation angle changes at the interval of the rotation angle Q when the coordinate conversion coefficient is acquired, the stored coordinate conversion coefficient may be used as it is. However, if the rotation angle of the stage 10 is continuously changed, the coordinate value converted by the interpolation calculation is calculated. In this interpolation, two sets of coordinate conversion coefficients corresponding to (n−1) · Q degrees and n · Q degrees sandwiching the input rotation angle are read, and two sets of coordinate conversion coefficients are used using the two sets of coordinate conversion coefficients. After calculating the coordinate values (X _A1 , Y _A1 , Z _A1 ), (X _A2 , Y _A2 , Z _A2 ), the difference between the input rotation angle with respect to (n−1) · Q degrees and n · Q degrees The final coordinate values (X _A , Y _A , Z _A ) may be calculated by a corresponding interpolation calculation. Further, two sets of coordinate transformation coefficients corresponding to the (n−1) · Q degree and the n · Q degree are set according to the difference of the input rotation angle with respect to the (n−1) · Q degree and the n · Q degree. To calculate one coordinate transformation coefficient, and calculate the final coordinate values (X _A , Y _A , Z _A ) by executing the above equation 12 using the coordinate transformation coefficient calculated by this interpolation computation. May be. If the input rotation angle is the reference rotation angle, that is, if the stage 10 is not rotated around the rotation axis A from the reference position, the coordinate conversion process in step S418 is not performed.

After the process of step S418, the data processing apparatus 40 is in a state in which the rotation angle around the rotation axis B of the stage 10 is at a reference angle (for example, 0 degree) from the coordinate-converted three-dimensional data in step S420. It is converted into the three-dimensional data of the object on the stage 10 acquired by the three-dimensional camera 20 in a state where the stage 10 is at the reference position. In this case, the coordinate conversion coefficient (matrix M _B ) for each minute predetermined angle Q acquired by rotating the stage 10 around the rotation axis B from the lower limit angle −Bmax to the upper limit angle Bmax and storing it in the nonvolatile memory. matrix value g and the matrix values a _{B in,} b B, from the _{c B),} the coordinate transformation coefficients (matrix M _B which corresponds to the rotation angle of the rotation axis B which is temporarily stored by input from the controller 30 The matrix value g and the matrix values a _B , b _B , c _B ) are selected. Then, coordinate values coordinate transformation by the process of the step _{S410 (X A, Y A,} Z A) and the matrix value g and the matrix values _a B of the selected coordinate transformation coefficients (matrix _{M B,} b _{B, c B} ) Is used to perform coordinate conversion by executing the following equation (13). The coordinate values (X _B , Y _B , Z _B ) in the following equation 13 are the objects on the stage 10 measured by the three-dimensional camera 20 in a state where the rotation angle around the rotation axis B of the stage 10 is at the reference angle. The three-dimensional data.

  Also in this case, when the rotation angle around the rotation axis B input from the controller 30 is continuous and there is an angle between every n · Q degrees, the interpolation calculation is performed as in step S418. The coordinate value converted by is calculated. Further, if the input rotation angle is the reference rotation angle, that is, if the stage 10 is not rotated around the rotation axis B from the reference position, the coordinate conversion process of step S420 is not performed.

After the process of step S420, the data processing apparatus 40 is in a state in which the rotation angle around the rotation axis C of the stage 10 is at a reference angle (for example, 0 degree) from the coordinate-converted three-dimensional data in step S422. It is converted into the three-dimensional data of the object on the stage 10 acquired by the three-dimensional camera 20 in a state where the stage 10 is at the reference position. In this case, which had been stored in the nonvolatile memory obtains rotated from 0 ° to stage 10 around the rotation axis C up to 360 °, fine coordinate transformation coefficients for each predetermined angle Q (matrix value of the matrix M _C g and matrix value a _C, b C, _c from among _C), the matrix value g of the coordinate transformation coefficients (matrix M _C corresponding to the rotation angle of the rotation axis C which is temporarily stored by input from the controller 30 And matrix values a _C , b _C , c _C ). Then, the coordinate values (X _B , Y _B , Z _B ) subjected to coordinate conversion by the process of step S420 are used using the selected coordinate conversion coefficients (matrix M _C and matrix values a _C , b _C , c _C ). Coordinate conversion is performed by executing the following equation (14). Note that the coordinate values (X _C , Y _C , Z _C ) in the following equation 14 are the objects on the stage 10 measured by the three-dimensional camera 20 in a state where the rotation angle around the rotation axis C of the stage 10 is at the reference angle. The three-dimensional data.

  Also in this case, when the rotation angle around the rotation axis C input from the controller 30 is continuous and there is an angle between n · Q degrees, as in the case of steps S418 and S420, The coordinate value converted by the interpolation operation is calculated. If the input rotation angle is a reference rotation angle, that is, if the stage 10 is not rotated around the rotation axis C from the reference position, the coordinate conversion process in step S422 is not performed.

  The reason why the coordinate transformation is performed in the order of the rotation axis A, the rotation axis B, and the rotation axis C by the processing in steps S418, S420, and S422 is that the rotation axis A of the stage 10 is as shown in FIG. The remaining two rotation axes B and C are fixed independently of each other. The rotation axis B is fixed if the rotation angle around the rotation axis A is fixed, and the rotation axis C is rotated around the remaining two rotation axes A and B. This is because if the angle is fixed, the angle is fixed. After the process of step S422, in step S424, the data processing device 40 stores a set of three-dimensional data obtained by coordinate conversion by the processes of steps S418, S420, and S422 in a nonvolatile memory. Thereby, in the nonvolatile memory, the three-dimensional data acquired by the three-dimensional camera 20 by rotating the stage 10 around the rotation axes A, B, and C, that is, the three-dimensional data of the desired coordinate system, A set of three-dimensional data converted into the three-dimensional data of the reference coordinate system corresponding to the reference position of the stage 10 is stored.

  After the process of step S424, the data processing device 40 determines whether or not the coordinate conversion coefficient of the camera coordinates is input from the controller 30 in step S426. As will be described later in detail regarding the input of the coordinate conversion coefficient of the camera coordinates, it is assumed that the input of the coordinate conversion coefficient of the camera coordinates is not input at this stage. Therefore, the data processing device 40 determines “No” in step S426 and proceeds to step S432. In step S432, the data processing apparatus determines whether there is an instruction to generate three-dimensional image data. If there is no instruction to generate the three-dimensional image data, the data processing device 40 determines “No” in step S432 and returns to step S402. The data processing apparatus 40 makes a “No” determination at step S402 until a measurement instruction for a new measurement object OB is received, and the process proceeds to step S426. Therefore, if there is no instruction to measure a new measurement object OB, input of a coordinate conversion coefficient of camera coordinates, and an instruction to generate three-dimensional image data, the data processing device 40 performs steps “S402, S426, and S432”. It is determined as “No”, and the circulation process including steps S402, S426, and S432 is continuously executed.

  On the other hand, when the operator operates the input device 31 again in this state and newly designates the rotational position of the stage 10 around the rotation axes A, B, and C, the controller 30 as described above. The motors 16, 17, and 18 are driven and controlled to rotate the stage 10 around the rotation axes A, B, and C to the rotation position designated by the operator. Thereafter, when the operator instructs the acquisition of the three-dimensional data of the measurement object OB on the stage 10 by operating the input device 31, the controller 30 determines “Yes” in step S <b> 302. Steps S304 to S308 are executed. On the other hand, the data processing device 40 determines “Yes” in step S402 in response to the measurement instruction of the object in the process of step S304 of the controller 30, and executes the process including steps S404 to S424 described above. To do. As a result, in the nonvolatile memory, the three-dimensional data acquired by the three-dimensional camera 20 by rotating the stage 10 around the rotation axes A, B, and C by a desired angle different from the above is used as the reference position of the stage 10. A new set of three-dimensional data converted into three-dimensional data in the corresponding coordinate system is stored. Then, if necessary, the operator rotates the stage 10 as many times as necessary around the rotation axes A, B, C, and causes the three-dimensional camera 20 to measure the measurement object OB. The set number of three-dimensional data can be stored in the nonvolatile memory.

  On the other hand, when the three-dimensional camera 20 is moved, for example, when the camera 20 that was at the position indicated by the broken line in FIG. 9 is moved to the position indicated by the solid line, it is true as described above before or after the movement of the camera 20. The balls 11a, 11b, and 11c are set on the stage 10. The movement of the camera 20 means that only the support device 21 is moved while the camera 20 is fixed to the support device 21, the case where only the camera 20 is rotated or displaced while the support device 21 is fixed, and the support device. This includes a case where the camera 20 is rotated or displaced while the lens 21 is moved. In this case, when the measurement object OB after the movement of the three-dimensional camera 20 is measured, the position of the measurement object OB with respect to the stage 10 is the position of the measurement object OB before the movement of the three-dimensional camera 20 with respect to the stage 10. If the measurement of the true spheres 11a, 11b, and 11c is possible without removing the measurement object OB from the stage 10, the measurement object OB is left on the stage 10. It is preferable. When the measurement object OB is removed from the stage 10, a mark or the like may be attached on the stage 10 so that the installation position of the measurement object OB with respect to the stage 10 can be reproduced. After the movement of the three-dimensional camera 20 and the setting of the true spheres 11a, 11b, and 11c, the operator inputs to the controller 30 that the camera 20 has been moved by operating the input device 31.

  In response to the input of the movement of the camera 20, the controller 30 determines “Yes” in step S310 of the object measurement program of FIG. 7, and performs the process of step S312 similar to step S102 of FIG. The rotation of the motors 16, 17, 18 is controlled so that the rotation angle around the rotation axes A, B, C of the stage 10 becomes the reference rotation angle (that is, 0 degree), that is, the stage 10 is set to the reference position. . After the processing in step S312, the controller 30 executes the above-described coordinate conversion coefficient acquisition routine of FIG. 6 in step S314. By the processing of steps S252 to S256 of this coordinate conversion coefficient acquisition routine, the center coordinates of the newly set true spheres 11a, 11b, 11c, that is, the X, Y, Z coordinate values (Xa1, Ya1, Za1) of the three fixed points, ( Xb1, Yb1, Zb1) and (Xc1, Yc1, Zc1) are calculated. Then, by the processing in step S258, the X, Y, Z coordinate values (Xa1, Ya1, Za1), (Xb1, Yb1, Zb1), (Xc1, Yc1, Zc1) of the three new fixed points and the above-mentioned FIG. X, Y, Z coordinate values (Xa0, Ya0, Za0) of three fixed points stored in the RAM or the non-volatile memory by the process of step S106 (step S208 in FIG. 5) when the stage 10 is at the reference position. , (Xb0, Yb0, Zb0) and (Xc0, Yc0, Zc0) are used to calculate the coordinate conversion coefficient.

In this case, the coordinate value of the three-dimensional data measured by the three-dimensional camera 20 after the movement is measured by the three-dimensional camera 20 arranged at the position where the coordinate conversion coefficient regarding the rotation angle about the rotation axes A, B, and C is acquired. A coordinate conversion coefficient for conversion to the coordinate value of the three-dimensional data (the coordinate value of the reference coordinate system) is calculated. That is, as shown in FIG. 9, the reference coordinate system in which the three-dimensional camera 20 is arranged at the position indicated by the broken line (the same position as in FIGS. 1 and 3) is the coordinate system S, and the three-dimensional camera 20 is at the position indicated by the solid line. If the arranged coordinate system is the coordinate system M, a coordinate conversion coefficient for converting the three-dimensional data of the coordinate system M into the three-dimensional data of the coordinate system S is calculated. Here, in order to distinguish from the above-described coordinate conversion coefficients related to the rotation angles around the rotation axes A, B and C, the matrix M will be referred to as the matrix M _S for the coordinate conversion coefficients related to the movement of the three-dimensional camera 20 hereinafter. In addition, the matrix values a, b, and c are set as matrix values a _S , b _S , and c _S. Then, after the process of step S258, the matrix value g and the matrix values a _S , b _S , c _S of the matrix M _S calculated in step S260 are stored in the nonvolatile memory, and the coordinate conversion coefficient is calculated in step S262. The execution of the acquisition routine is terminated.

Returning to the description of the object measurement program in FIG. 7 again, after the process of step S314, the controller 30 calculates the matrix value g and the matrix value of the matrix M _S calculated and stored in the nonvolatile memory in step S316. The coordinate transformation coefficient of the camera coordinate system composed of a _S , b _S , and c _S is output to the data processing device 40.

In response to the output of the coordinate conversion coefficient of the camera coordinate system, the data processing device 40 determines “Yes” in step S426, that is, the input of the coordinate conversion coefficient of the camera coordinate, and performs the processing of steps S428 and S430. Run. In step S428, the coordinate transformation coefficient of the camera coordinate system composed of the matrix value g of the matrix M _S and the matrix values a _S , b _S , c _S output from the controller 30 is stored in the nonvolatile memory. In step S430, the camera movement flag CMF is set to “1”.

  In this state, the operator removes the true spheres 11a, 11b, and 11c on the stage 10. When the measurement object OB is detached from the stage 10 when the true spheres 11a, 11b, and 11c are attached to the stage 10, the measurement object OB is measured before the three-dimensional camera 20 is moved. Is installed at the same position on the stage 10. When the operator operates the input device 31 to instruct acquisition of the three-dimensional data of the measurement object OB on the stage 10, the controller 30 determines “Yes” in step S <b> 302. Steps S304 to S308 are executed. Even if the rotation position of the stage 10 around the rotation axes A, B, and C is changed by the operation of the input device 31 before the instruction to acquire the three-dimensional data of the measurement object OB, as described above. Good. On the other hand, the data processing device 40 determines “Yes” in step S402 in response to the measurement instruction of the measurement object OB by the processing of step S304 of the controller 30, and performs the above-described processing of steps S404 to S412. The rotation angle of the stage 10 about the rotation axes A, B, and C and the three-dimensional data newly acquired by the three-dimensional camera 20 are stored in the nonvolatile memory.

After the processing in step S412, the controller 30 determines in step S414 whether the camera movement flag CMF is “1”. In this case, since the camera movement flag CMF is set to “1” by the process of step S430 described above, the controller 30 determines “Yes” in step S414, and uses the newly acquired three-dimensional data. Convert coordinates. In this coordinate transformation, the newly acquired three-dimensional data is calculated by using the matrix value g and matrix values a _S , b _S , and c _S of the matrix M _S stored in the nonvolatile memory. The coordinates are converted by executing. Note that the coordinate values (X _M , Y _M , Z _M ) in the following formula 15 are the three-dimensional data of the coordinate system M in which the three-dimensional camera 20 is at the solid line position in FIG. Further, the coordinate values (X _S , Y _S , Z _S ) in the following formula 15 are the three-dimensional data of the coordinate system S in which the three-dimensional camera 20 is at the position of the broken line in FIG.

  After the process of step S416, the controller 30 obtains the three-dimensional data obtained by converting the three-dimensional data of the coordinate system M into the coordinate system S at the reference position of the stage 10 in the coordinate system S by the processes of steps S418 to S422 described above. It is converted into a set of three-dimensional data in a corresponding reference coordinate system. Then, the set of converted three-dimensional data is stored in the nonvolatile memory by the process of step S424.

  In this state, when the operator operates the input device 31 again to newly indicate the rotational positions around the rotation axes A, B, and C of the stage 10, the controller 30, as described above, The motors 16, 17, and 18 are driven and controlled to rotate the stage 10 around the rotation axes A, B, and C to the rotation position designated by the operator. Thereafter, when the operator instructs the acquisition of the three-dimensional data of the measurement object OB on the stage 10 by operating the input device 31, the controller 30 determines “Yes” in step S <b> 302. Steps S304 to S308 are executed. On the other hand, the data processing device 40 determines “Yes” in step S402 in response to the measurement instruction of the measurement object OB by the process of step S304 of the controller 30, and the process including steps S404 to S424 described above. Execute. As a result, the non-volatile memory stores the three-dimensional data acquired by the three-dimensional camera 20 by rotating the stage 10 after the movement of the three-dimensional camera 20 around the rotation axes A, B, C by a desired angle different from the above. A new set of three-dimensional data obtained by converting the above into three-dimensional data in the reference coordinate system is stored. In this case as well, the operator rotates the stage 10 around the rotation axes A, B, and C as many times as necessary, and causes the three-dimensional camera 20 to measure the measurement object OB. The number of sets of three-dimensional data corresponding to the number of times can be stored in the nonvolatile memory.

  Further, when it is necessary to move the three-dimensional camera 20 and measure the measurement object OB three-dimensionally, the true spheres 11a, 11b, and 11c are set on the stage 10 as described above, and FIG. The coordinate transformation coefficient of the camera coordinate system accompanying the movement of the three-dimensional camera 20 is acquired by the processes of steps S310 to S316 and steps S426 to S430 of FIG. Then, the measurement object OB is measured by the three-dimensional camera 20 by the processing of steps S302 to S308 in FIG. 7 and steps S402 to S424 in FIG. 8, and the three-dimensional data obtained by the measurement is converted into the three-dimensional data in the reference coordinate system. The three-dimensional data converted into the reference coordinate system is newly stored in the nonvolatile memory. Also in this case, the stage 10 is rotated around the rotation axes A, B, and C to obtain three-dimensional data of the reference coordinate system. As a result of such processing, a plurality of sets of three-dimensional data relating to the measurement object OB acquired while rotating the stage 10 about the rotation axes A, B, and C for each of a plurality of positions of the three-dimensional camera 20 are stored in a nonvolatile memory. Will be remembered. A plurality of sets of three-dimensional data stored in the nonvolatile memory are all converted to data in the reference coordinate system.

  Thus, after acquiring the three-dimensional data of the measurement object OB while rotating the stage 10 around the rotation axes A, B, and C for each of a plurality of positions of the three-dimensional camera 20, the operator can input the input device 31. To confirm the end of the measurement in order to confirm the three-dimensional image of the measurement object OB. In response to this measurement end instruction, the controller 30 determines “Yes” in step S318 of the object measurement program of FIG. 7, and generates the three-dimensional image data in the data processing device 40 in step S320. Instructing, the execution of the object measurement program is terminated in step S322.

  In response to the instruction to generate the three-dimensional image data, the data processing device 40 determines “Yes” in step S432, and executes the processes in steps S434 to 438. In step S434, three-dimensional data of a plurality of sets of measurement objects OB stored in the nonvolatile memory is synthesized. In this synthesis, a plurality of sets of three-dimensional data in the reference coordinate system are aggregated into a set of three-dimensional data. In step S436, 3D image data representing the 3D shape of the measurement object OB is generated based on the combined 3D data, and the generated 3D image data is stored in the nonvolatile memory. In step S438, the three-dimensional image data stored in the nonvolatile memory is supplied to the display device 41. The display device 41 displays a three-dimensional image of the measurement object OB using the supplied three-dimensional image data.

  In the embodiment operating as described above, the three-dimensional camera of the coordinate system M is obtained by using the coordinates of the fixed point of the stage 10 stored by the process of step S106 of FIG. 4A by the process of steps S310 to S314 of FIG. A coordinate conversion coefficient for converting the three-dimensional data by 20 into the three-dimensional data by the three-dimensional camera 20 in the coordinate system S (reference coordinate system) is acquired. 8, the three-dimensional data from the three-dimensional camera 20 in the coordinate system M is converted into the three-dimensional data from the three-dimensional camera 20 in the coordinate system S using the coordinate conversion coefficient. Then, the three-dimensional data converted into the coordinate system S is converted into the three-dimensional data of the reference coordinate system corresponding to the reference position of the stage 10 by the processing of steps S418 to S422 in FIG. Therefore, even if the position of the three-dimensional camera 20 with respect to the stage 10 is changed, a coordinate conversion coefficient for each rotation angle can be obtained by performing only one measurement for obtaining the coordinates of the fixed point defined in the stage 10, The measurement efficiency is kept good.

  Further, the fixed points defined in the stage 10 are obtained by mounting the true spheres 11 a, 11 b, 11 c that can define fixed points at fixed positions with respect to the stage 10 in the holes 10 a, 10 b, 10 c of the stage 10. Therefore, the true spheres 11 a, 11 b, and 11 c need only be attached to the stage 10 only when performing measurements for obtaining the coordinates of the fixed points defined on the stage 10, so that the measurement object OB is placed on the stage 10. It does not get in the way.

  Although one embodiment of the present invention has been described above, the present invention is not limited to the above-described embodiment, and various modifications can be made without departing from the object of the present invention.

In the above-described embodiment, the coordinate conversion coefficient relating to the rotation of the stage 10 and the coordinate conversion relating to the movement of the three-dimensional camera 20 are performed by the processing in steps S116, S126, S138, S148 in FIG. 4A, step S160 in FIG. In obtaining the coefficients, coordinate conversion coefficients (matrix value g of matrix M _A and matrix values a _A , b _A , c _A , matrix value g and the matrix value _a B matrix _{M B, b B, c B} , matrix values g and matrix value _a C matrix _{M C, b C, c C} , as well as the matrix of the matrix _{M S} value g and the matrix value a _S , b _S , c _S ) are calculated. However, instead of this, two or more fixed points and one or more vectors may be defined, or one or more fixed points and two or more vectors may be defined to calculate the coordinate conversion coefficient. In order to calculate a coordinate conversion coefficient for converting the coordinate value of the first coordinate system to the coordinate of the second coordinate system, for the fixed point, the fixed point coordinate value of the first coordinate system is the coordinate value (x ′, y ′). , Z ′), and if the fixed point coordinate value of the second coordinate system is the coordinate value (x, y, z), these coordinate values are substituted into Equation 16 similar to Equations 2 and 3 above. As for the vector, if the vector component of the first coordinate system is (α ′, β ′, γ ′) and the vector component of the second coordinate system is (α, β, γ), these The vector component is substituted into Equation 17 similar to Equation 8.

If the vector of the first coordinate system and the vector of the second coordinate system are the same, the calculation of Equation 17 may be applied, but the vector of the first coordinate system and the vector of the second coordinate system. When the sizes of are different, the coefficient K is calculated by executing the calculation of the following equation (18). Then, the calculated coefficient K is applied to the vector of the second coordinate system and substituted into the equation (17). When two or more fixed points and one or more vectors are defined by substituting into the equations 16 and 17, two or more simultaneous equations and a number related to the equation 16 composed of three matrix values g are obtained. One or more simultaneous equations for 17 will be prepared. In addition, when one or more fixed points and two or more vectors are defined, one or more simultaneous equations relating to Equation 16 and three or more simultaneous equations relating to Equation 17 are prepared. Will be. In any case, three or more simultaneous equations composed of three matrix values g are prepared, and the coordinate transformation coefficients (g ₁₁ , g ₁₂ , g ₁₃ , g ₂₁ , g ₂₂ , g ₂₃ are obtained by solving the simultaneous equations. , G ₃₁ , g ₃₂ , g ₃₃ , a, b, c) are calculated.

  Next, a specific method regarding the definition of two or more fixed points and one or more vectors and the definition of one or more fixed points and two or more vectors will be described. When defining two or more fixed points and one or more vectors, for example, as shown in FIG. 10 (A), two central spheres 11a and 11b set on the stage 10 are represented by two central coordinates. While defining as a fixed point, the normal vector on the upper surface of the stage 10 may be defined as one vector.

  Further, when defining one or more fixed points and two or more vectors, for example, as shown in FIG. 10B, the central coordinates of one true sphere 11a set on the stage 10 are set as one fixed point. And a vector defined by the direction of the linear pattern 10e created on the stage 10 and a normal vector on the upper surface of the stage 10 may be defined as two vectors. Further, as shown in FIG. 10C, two vectors are defined by the directions of the two linear intersecting patterns 10f and 10g created on the stage 10, and the two linear patterns 10f and 10g are defined. You may make it define a fixed point with an intersection. Further, as shown in FIG. 10D, two vectors are defined by the normal vector of the upper surface of the stage 10 and the direction of the linear pattern 10h created on the stage, and the pattern 10h and the pattern 10i You may make it define a fixed point by an intersection. Further, as shown in FIGS. 10E and 10F, the stage 10 is configured to have at least one flat side surface and apex (corner) in addition to the upper surface, and the normal to the upper surface. The vector and the side normal vector may be defined as two vectors, and the vertex may be defined as a fixed point.

  When the linear patterns 10e to 10i described above are used, the reflectance of the patterns 10e to 10i is different from the reflectance of other portions on the upper surface of the stage 10, and the three-dimensional shape is measured by the three-dimensional camera 20. If the amount of reflected light is also measured at the same time, three-dimensional data regarding the patterns 10e to 10i can be extracted. In addition, when using normal vectors of the upper surface and side surface of the stage 10, the reflectance is different between the upper surface and side surface of the stage 10, and reflection is performed when measuring the three-dimensional shape by the three-dimensional camera 20. If the amount of light is also measured at the same time, three-dimensional data relating to the upper and side surfaces of the stage 10 can be extracted separately. A method for calculating fixed point coordinates and normal vector components from the three-dimensional data obtained by these measurements is a conventionally known technique (for example, Japanese Patent Laid-Open No. 2005-249402, Japanese Patent Laid-Open No. 2005-352630). And Japanese Patent Application Laid-Open No. 2008-164493). In this way, by using the normal vectors of the vertices, the top surface, and the side surface in the shape of the stage 10 and the points or straight line portions determined from the pattern drawn on the stage 10, the fixed points or vector components defined in the stage 10 are obtained. Since it is sufficient to measure the stage 10 as it is, the efficiency is high.

  Further, in the acquisition of coordinate transformation coefficients by the processing of steps S116, S126, S138, S148 of FIG. 4A, step S160 of FIG. 4B, and step S314 of FIG. 7, stage 10 is obtained by the processing of the reference fixed point coordinate acquisition routine of FIG. The coordinates of a fixed point in a state where is at the reference position are stored, and the stored coordinates of the fixed point are used. Therefore, when a fixed point and a vector are used for coordinate conversion as in the modified example, instead of storing the coordinates of the fixed point in the reference fixed point coordinate acquisition routine of FIG. The coordinates of the fixed point and the vector component used in the acquisition of the state where the stage 10 is at the reference position are acquired and stored in the RAM or the nonvolatile memory.

  Various methods other than the example shown in FIG. 10 can be considered as a method for defining fixed points and vectors. For example, when a fixed point is defined using an object, it is preferable to use a true sphere because data processing for calculating fixed point coordinates from three-dimensional data is the easiest. However, as long as a fixed point can be defined, an object having a shape such as a cylinder, a cone, a pyramid, or a rectangular parallelepiped may be used, and a pattern such as a circle or a polygon may be used.

  Moreover, in the said embodiment, three-dimensional shape measurement was performed for every rotation angle in each rotating shaft, the fixed point coordinate was acquired, and the coordinate conversion coefficient was acquired for every rotation angle. However, instead of this, as shown in FIG. 11, one fixed point is defined and fixed point coordinates are respectively set at three or more rotation angles while rotating the stage 10 around the three rotation axes A, B, and C. And the positions of the rotation axes A, B, and C are calculated from the plane determined by the fixed point coordinates and the center coordinates of the circle determined by the fixed point coordinates, respectively, and theoretically the positions of the rotation axes A, B, and C are calculated. The coordinate conversion coefficient for each rotation angle may be acquired as a function of the rotation angle. In this case as well, various shapes of objects and patterns can be used to define the fixed point, but it is preferable to use a true sphere 11a set on the stage 10 as shown in FIG.

  The coordinate conversion coefficient for each rotation angle on each of the rotation axes A, B, and C is obtained by rotating the coordinate axes of the reference coordinate system (camera coordinate system) at the reference position of the stage 10 around the respective coordinate axes, Assuming a coordinate axis in which one intersects perpendicularly and one of the coordinate axes is parallel to the rotation axis, and this coordinate axis rotates around the rotation axis, it can be calculated as a function of the rotation angle around the rotation axis. it can. Specifically, the coordinate conversion coefficient may be obtained by calculating for each of the rotation axes A, B, and C in the following order.

(1) A plane H perpendicular to the rotation axis and including the origin O of the reference coordinate system (camera coordinate system) is calculated.
(2) The angle formed between each coordinate axis and the plane H is calculated, and a straight line vector J formed from the plane I and the plane H including the origin O and perpendicular to the coordinate axis a having the smallest angle with the plane H is calculated. Then, the angle between the vector J and two coordinate axes other than the coordinate axis a is calculated, and the coordinate axis b, which is another coordinate axis, is rotated within the plane H by rotating the coordinate axis a by the smaller angle. The coefficient M1 is calculated. The coordinate conversion coefficient M1 and coordinate conversion coefficients M2 and M3, which will be described later, are 3 × 3 matrices as shown in Equation 3 above.
(3) The coordinate transformation coefficient M2 is calculated by rotating the coordinate axis b by an angle formed by the plane H and the coordinate axis a to bring the coordinate axis a into the plane H.

(4) The intersection point K between the plane H and the rotation axis is calculated, the angle between the vector J from the origin O toward the intersection point K and the vector J is calculated, the coordinate axis c other than the coordinate axes a and b is rotated by this angle, and the coordinate axis A coordinate conversion coefficient M3 for calculating the direction of a from the origin O to the intersection K direction is calculated.
(5) Coordinate conversion coefficients M (θ) and T (θ) are calculated when the coordinate axis formed in (4) is rotated about the rotation axis by the rotation angle θ. In this case, since the origin is moved, the coordinate conversion coefficients are M (θ) which is a 3 × 3 matrix and T (θ) which is a 3 × 1 matrix (vector component). Both the coordinate conversion coefficients M (θ) and T (θ) can be calculated as a function of the rotation angle θ.
(6) Coordinate transformation coefficients M3 ⁻¹ , M2 ⁻¹ , M1 ⁻¹ for rotating the coordinate axes rotated in the above (5) opposite to the rotations performed in the above (2), (3), (4) Calculate

(7) Coordinate conversion coefficient Mt (θ) = M1 ⁻¹ · M2 ⁻¹ · M3 ⁻¹ · M (θ) · M3 · M2 · M1 and coordinates when rotating in the order of (2) to (6) Conversion coefficient Tt (θ) = M1 ⁻¹ · M2 ⁻¹ · M3 ⁻¹ · T (θ) is calculated.
(8) The coordinate transformation coefficients Mt (θ) and Tt (θ) obtained in the above (7) rotate the coordinate axis of the reference coordinate system (camera coordinate system) at the reference position of the stage 10 by the rotation angle θ around the rotation axis. Is the coordinate conversion coefficient. The coordinate conversion coefficient for making the coordinate axis after being rotated by the rotation angle θ be the coordinate axis of the reference coordinate system (camera coordinate system) at the reference position of the stage 10 is Mt (, which is an inverse matrix of the coordinate conversion coefficient Mt (θ). θ) ⁻¹ and −Mt (θ) ⁻¹ · Tt (θ).

  This type of technology is also a conventionally known technology (see Japanese translations of PCT publication No. 2000-509505 and Japanese Patent Application Laid-Open No. 2008-32449).

  In the embodiment, the stage 10 has the rotation axis A fixed, the rotation axis B is fixed when the rotation angle of the rotation axis A is fixed, and the rotation axis C is fixed when the rotation axes A and B are fixed. Although fixed, the relationship between the rotation axes A, B, and C may be arbitrary as long as coordinate conversion is performed from coordinate conversion using a fixed rotation axis. In the above embodiment, there are three rotation axes. However, if coordinate conversion is performed from coordinate conversion using fixed rotation axes, the number of rotation axes may be two, or four or more. May be.

  Further, in the above embodiment, when obtaining a coordinate conversion coefficient for converting the three-dimensional data of the coordinate system M into the three-dimensional data of the coordinate system S (reference coordinate system), the rotation axes A, B, and C of the stage 10 are obtained. The fixed point coordinates were obtained by measuring the three-dimensional shapes of the true spheres 11a, 11b, and 11c with the rotation angle as the reference angle (0 degree). It is possible to obtain a conversion factor. In this case, the coordinate system S is stored in the non-volatile memory by executing the coordinate conversion coefficient setting program of FIGS. 4A and 4B and uses the coordinate conversion coefficients at the rotation angles around the rotation axes A, B, and C of the stage 10. A fixed point coordinate in the (reference coordinate system) is converted into a fixed point coordinate in the coordinate system based on the current rotation angle by reverse coordinate conversion, and the fixed point coordinate obtained by measuring the fixed point coordinate and the three-dimensional shape is used. What is necessary is just to acquire the coordinate transformation coefficient which carries out the coordinate transformation of the three-dimensional data to the three-dimensional data of the coordinate system S (reference coordinate system).

  Furthermore, in the above-described embodiment, the three-dimensional camera 20 performs measurement by irradiating laser light. However, any method can be used as long as the three-dimensional shape of the measurement object OB can be measured. May be. For example, a method of measuring by projecting a lattice pattern, a method of measuring using ultrasonic waves, or a method of measuring by contacting a probe may be used.

DESCRIPTION OF SYMBOLS 10 ... Stage, 11a, 11b, 11c ... True sphere, 16, 17, 18 ... Motor, 20 ... Three-dimensional camera, 30 ... Controller, 31 ... Input device, 40 ... Data processing device, 41 ... Display device, OB ... Measurement Object

Claims (10)

A stage that can be rotated about two or more rotation axes and that can define a fixed point or vector for calculating a coordinate transformation coefficient;
A three-dimensional camera for measuring a three-dimensional shape of an object placed on the stage;
A processing device for processing three-dimensional data representing a three-dimensional shape measured by the three-dimensional camera,
In the processing device,
A fixed point or vector of the stage acquired by measurement of the three-dimensional camera arranged at the first camera position is used to represent a three-dimensional shape measured by the three-dimensional camera with the stage rotated from a reference position. The two or more rotation axes represent a rotation coordinate conversion coefficient for converting three-dimensional data into three-dimensional data of a reference coordinate system representing a three-dimensional shape measured by the three-dimensional camera in a state where the stage is at a reference position. Rotating coordinate conversion coefficient acquisition means for acquiring and storing for each of a plurality of rotation angles for each of,
The three-dimensional data representing the three-dimensional shape of the measurement object measured by the three-dimensional camera with the stage rotated from the reference position is converted into the three-dimensional data of the reference coordinate system using the rotation coordinate conversion coefficient. A rotating coordinate conversion means for converting,
In the three-dimensional shape measuring apparatus that combines the three-dimensional data of the measurement object measured from different viewpoints by rotating the stage with the three-dimensional data of the reference coordinate system,
A fixed point / vector storage means for storing the fixed point coordinates or vector components of the stage acquired by the measurement of the three-dimensional camera arranged at the first camera position in the processing device;
The fixed point coordinates or vector components of the stage acquired by measurement of the three-dimensional camera arranged at a second camera position different from the first camera position, and the stage stored in the fixed point / vector storage means Three-dimensional data representing a three-dimensional shape measured by the three-dimensional camera at the second camera position is measured by the three-dimensional camera at the first camera position using a fixed point coordinate or a vector component. Camera coordinate conversion coefficient acquisition means for acquiring a camera coordinate conversion coefficient for converting coordinates into three-dimensional data representing a three-dimensional shape;
Three-dimensional data representing a three-dimensional shape of a measurement object measured by the three-dimensional camera at the second camera position is measured by the three-dimensional camera at the first camera position using the camera coordinate conversion coefficient. A three-dimensional shape measuring apparatus, comprising: camera coordinate conversion means for converting into three-dimensional data representing the three-dimensional shape of the measured object.   2. The fixed point / vector storage means stores coordinates or vector components of a fixed point of the stage in conjunction with acquisition of a rotation coordinate conversion coefficient by the rotation coordinate conversion coefficient acquisition means. Three-dimensional shape measuring device. Fixed point or vector as defined above stage, claim 1, characterized in that it is obtained by mounting the can body defining a vector to the stage in fixed point or a certain direction in a fixed position relative to the stage Or the three-dimensional shape measuring apparatus of 2. The three-dimensional shape measuring apparatus according to claim 1 , wherein the fixed point or vector defined in the stage is a normal vector of a vertex or a plane in the shape of the stage. The three-dimensional shape measuring apparatus according to claim 1 , wherein the fixed point or vector defined in the stage is a point or a straight line portion determined from a pattern drawn on the stage. A stage capable of rotating around two or more rotation axes and capable of defining a fixed point or a vector for calculating a coordinate transformation coefficient; a three-dimensional camera for measuring a three-dimensional shape of an object placed on the stage; Applied to a three-dimensional shape measuring apparatus comprising:
A fixed point or vector of the stage acquired by measurement of the three-dimensional camera arranged at the first camera position is used to represent a three-dimensional shape measured by the three-dimensional camera with the stage rotated from a reference position. The two or more rotation axes represent a rotation coordinate conversion coefficient for converting three-dimensional data into three-dimensional data of a reference coordinate system representing a three-dimensional shape measured by the three-dimensional camera in a state where the stage is at a reference position. Are acquired and stored for each of a plurality of rotation angles,
The three-dimensional data representing the three-dimensional shape of the measurement object measured by the three-dimensional camera with the stage rotated from the reference position is converted into the three-dimensional data of the reference coordinate system using the rotation coordinate conversion coefficient. In the three-dimensional shape measurement method of combining the three-dimensional data of the measurement object measured from different viewpoints by converting and rotating the stage, with the three-dimensional data of the reference coordinate system,
Store the coordinates or vector components of the fixed point of the stage acquired by the measurement of the three-dimensional camera arranged at the first camera position,
The fixed point coordinates or vector components of the stage acquired by the measurement of the three-dimensional camera arranged at a second camera position different from the first camera position, and the stored fixed point coordinates or vector of the stage And the three-dimensional data representing the three-dimensional shape measured by the three-dimensional camera at the second camera position, and the three-dimensional shape measured by the three-dimensional camera at the first camera position. Obtain the camera coordinate conversion coefficient for coordinate conversion to 3D data,
Three-dimensional data representing a three-dimensional shape of a measurement object measured by the three-dimensional camera at the second camera position is measured by the three-dimensional camera at the first camera position using the camera coordinate conversion coefficient. A three-dimensional shape measuring method, wherein the three-dimensional data representing the three-dimensional shape of the measured object is converted. 7. The three-dimensional shape measuring apparatus according to claim 6 , wherein the storage of the fixed point coordinates or vector components of the stage is performed in conjunction with the acquisition of the rotational coordinate conversion coefficient. 6. fixed point or vector as defined above stage, characterized by being obtained by mounting the can body defining a vector to the stage in fixed point or a certain direction in a fixed position relative to the stage Or the three-dimensional shape measuring method of 7 . The three-dimensional shape measuring method according to claim 6 or 7 , wherein the fixed point or vector defined in the stage is a normal vector of a vertex or a plane in the shape of the stage. The three-dimensional shape measuring method according to claim 6 or 7 , wherein the fixed point or vector defined in the stage is a point or a straight line portion determined from a pattern drawn on the stage.

Images