JP2015203675A - Image processing apparatus, image processing system, three-dimensional measuring instrument, image processing method, and image processing program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To remove noise from a measurement result, to improve precision of three-dimensional point group data for a three-dimensional model to be formed.SOLUTION: An image processing apparatus acquires a measurement result of an object arranged in a three-dimensional space, as first point group data acquired from a first direction and second point group data acquired from a second direction different from the first direction, which are aggregates of point data. When point data indicating a first position of the object for first point data included in the first point group data is not included in the second point group data, the first point data is deleted from the first point group data. When point data indicating a second position of the object for second point data included in the second point group data is not included in the first point group data, the second point data is deleted from the second point group data.

Description

  The present invention relates to an image processing apparatus, an image processing system, a three-dimensional measuring instrument, an image processing method, and an image processing program.

  Patent Document 1 discloses a point sequence generation method for generating a plurality of point sequences for defining a reproduction curved surface of an object from a point group consisting of three-dimensional coordinate information representing the curved surface of the object, , A point sequence extraction process that extracts a point sequence consisting of multiple points in the point cloud, an analysis point sequence creation process that creates an analysis point sequence from the point sequence, and a curved surface analysis using the analysis point sequence , The surface analysis process to calculate the surface information, and at each point of each analysis point sequence, detect the feature point where the direction of the analysis point sequence and the direction of the normal section including the point where the normal curvature is zero Point sequence that performs a feature point detection process to be performed, and a point sequence generation process in which feature points are inserted into a plurality of point sequences in a point cloud composed of analysis point sequences or three-dimensional coordinate information to form new point sequences A generation method is disclosed.

JP 2011-180787 A

  The invention described in Patent Document 1 defines a curved surface with high accuracy from a point cloud, but does not consider whether the data of the point cloud is correct. Therefore, there is a problem that a curved surface cannot be defined with high accuracy when the point cloud includes incorrect data.

  14A and 14B are diagrams for explaining a problem when a point cloud includes incorrect data. FIG. 14A shows a measurement object, and FIG. 14B shows a result of three-dimensional measurement of the measurement object (point cloud). ), And (C) shows a result of incorrect shape recognition.

  When the cross section of the L-shaped steel 300 as shown in FIG. 14 (A) is measured with a three-dimensional measuring instrument, an L-shaped point group 301 should be obtained as shown in FIG. 14 (B). is there. However, as shown in FIG. 14C, a point group 301A including points formed at positions where there is no L-shaped steel may be obtained by irregular reflection or the like. In such a case, if a three-dimensional model is to be generated based on the point cloud 301, there is a possibility that it is not recognized as a cylindrical object 302 as shown in FIG. is there.

  Therefore, the present invention removes noise from the measurement result and increases the accuracy of the three-dimensional point cloud data that is the basis of the generated three-dimensional model, an image processing apparatus, an image processing system, a three-dimensional measuring instrument, an image A processing method and an image processing program are provided.

  In order to solve the above problems, an image processing apparatus according to the present invention, for example, a point cloud data acquisition unit that acquires measurement results of an object provided in a three-dimensional space as point cloud data that is an aggregate of point data. A point cloud data acquisition unit that acquires first point cloud data acquired from a first direction and second point cloud data acquired from a second direction different from the first direction; When the point data obtained by measuring the first position of the object that is the basis of the first point data included in the first point group data is not included in the second point group data, The point data obtained by deleting the first point data from the first point cloud data and measuring the second position of the object that is the basis of the second point data included in the second point cloud data is If not included in the first point cloud data, the second point data is converted to the second point data. Comprising a point group data deletion unit for deleting from the data, and an output unit for outputting a first point group data and the second point group data after being removed by the point group data deletion unit.

  According to the present invention, noise can be removed from the measurement result, and the accuracy of the three-dimensional point cloud data that is the basis of the generated three-dimensional model can be increased.

It is a figure which shows the structural example of the image processing apparatus 1 which concerns on embodiment of this invention. It is an example of the whole schematic diagram of the substation which uses the image processing apparatus. 2 is a diagram illustrating a hardware configuration of the image processing apparatus 1. FIG. 3 is a flowchart showing a processing flow of the image processing apparatus 1. It is an example which shows a mode that the three-dimensional measuring device 20 measures the switch 202 which is plant facilities. The position of the three-dimensional measuring device 20, the arbitrary point P included in the three-dimensional point group data 21 (first three-dimensional point group data 211, second three-dimensional point group data 212), and the error region R It is a figure which shows a relationship. It is a figure explaining the method in which the point data generation part 112 produces | generates several point data P1. It is a figure which shows the H-shaped steel which is an example of a measuring object. It is a figure explaining the matching in the case of measuring the end face of H type steel. It is a figure explaining the matching in the case of measuring the side of H-shaped steel. It is a figure explaining point data deletion. It is a display example. It is a display example. It is a figure explaining a problem when a point cloud contains incorrect data.

  Hereinafter, examples will be described with reference to the drawings.

  [Description of Configuration] FIG. 1 is a diagram showing a configuration example of an image processing apparatus 1 according to an embodiment of the present invention. The image processing apparatus 1 mainly includes a control unit 11, an output unit 12, an input unit 13, a communication unit 14, and a storage unit 15. The image processing apparatus 1 is mainly connected to a three-dimensional measuring device 20, a display device 30, and a three-dimensional CAD server (not shown) by wire or wirelessly.

  Note that the image processing device 1, the three-dimensional measuring device 20, and the display device 30 may be in the same or close locations, or different locations (for example, the three-dimensional measuring device 20 and the display device 30 are on-site, the image processing device). 1 may be in the office.

  The control unit 11 performs processing for controlling the entire image processing apparatus 1. The internal configuration of the control unit 11 will be described in detail later.

  The output unit 12 outputs a program execution state, a processing result, and the like.

  The input unit 13 inputs a program execution start instruction and a stop instruction from the user.

  The communication unit 14 exchanges data between the units of the image processing apparatus 1. In addition, the communication unit 14 communicates with other devices (for example, the three-dimensional measuring device 20 and a three-dimensional CAD server (not shown)) to exchange data.

  The storage unit 15 temporarily stores the acquired various data and stores the created data.

  The three-dimensional measuring device 20 irradiates each part of plant equipment (target object) in a plant such as a power plant 200 (described later) to obtain a three-dimensional measurement result. Since the three-dimensional measuring device 20 is already common, the description thereof is omitted.

  FIG. 2 is an overall view of the power plant 200. The power plant 200 includes, for example, a metal clad 201 having a size of several tens to several hundreds of meters, a switch 202, and the like. The three-dimensional measuring device 20 irradiates a laser beam to a plant facility such as the metal cladding 201 and the switch 202 from a distance of about 100 m, for example, and acquires a three-dimensional measurement result.

  The measurement result of the object by the three-dimensional measuring device 20 is processed by a processing device (not shown) or the like and is output to the image processing device 1 as three-dimensional point cloud data 21. The three-dimensional point group data 21 is a collection of point data. In the present embodiment, the three-dimensional measuring instrument 20 outputs a result of irradiating the plant equipment with a laser beam from a plurality of directions (for example, two directions), that is, a plurality of three-dimensional point cloud data 21 to the image processing apparatus 1. Is done.

  Further, three-dimensional CAD data 22 is output to the image processing apparatus 1 from a three-dimensional CAD server (not shown). The three-dimensional CAD data 22 is three-dimensional CAD data that is three-dimensional design data such as plant equipment or catalog data. In the present embodiment, the three-dimensional design data is polygon data, but is not limited to this.

  Returning to the description of FIG. The display device 30 includes a display device such as a CRT monitor or a liquid crystal monitor.

  Next, each component of the control unit 11 will be described in detail. As shown in FIG. 1, the control unit 11 mainly includes a point cloud data acquisition unit 111, a point data generation unit 112, a point cloud data matching unit 113, a point data deletion unit 114, and a three-dimensional CAD data acquisition unit. 115, a three-dimensional CAD data matching unit 116, and a three-dimensional CAD data generation unit 117.

  The point cloud data acquisition unit 111 acquires a plurality of 3D point cloud data 21 output from the 3D measuring device 20.

  The point data generation unit 112 generates new point data around the point data for each point data included in the three-dimensional point group data 21. The processing performed by the point data generation unit 112 will be described in detail later.

  The point cloud data matching unit 113 matches (superimposes) the plurality of three-dimensional point cloud data 21 after the point data generation unit 112 has processed. The processing performed by the point cloud data matching unit 113 will be described in detail later.

  The point data deletion unit 114 deletes point data that is not included in all of the plurality of three-dimensional point group data 21 from each of the three-dimensional point group data 21 as a result of matching by the point group data matching unit 113. When the point data deletion unit 114 is matched with the three-dimensional CAD data 22 by the three-dimensional CAD data matching unit 116, the point data matched with the three-dimensional CAD data 22 is obtained from the three-dimensional point cloud data 21. Do not delete. The processing performed by the point data deletion unit 114 will be described in detail later.

  The three-dimensional CAD data acquisition unit 115 acquires the three-dimensional CAD data 22 from a three-dimensional CAD server or the like (not shown).

  When the 3D CAD data 22 is acquired, the 3D CAD data matching unit 116 performs the 3D CAD data 22 on the 3D point cloud data 21 after the point data generation unit 112 generates the point data. And matching (superposition). The processing performed by the three-dimensional CAD data matching unit 116 will be described in detail later.

  In the present embodiment, the three-dimensional CAD data acquisition unit 115 acquires the three-dimensional CAD data 22, and the three-dimensional CAD data matching unit 116 matches the three-dimensional CAD data 22 with the three-dimensional point cloud data 21. However, the three-dimensional CAD data 22 is an example of three-dimensional data (data that can be recognized as a three-dimensional figure), and is not limited to three-dimensional CAD data.

  The three-dimensional CAD data generation unit 117 generates three-dimensional CAD data based on the three-dimensional point cloud data 21 from which the point data has been deleted by the point data deletion unit 114. Since a method for generating three-dimensional CAD data based on the three-dimensional point cloud data 21 is known, a description thereof will be omitted. The three-dimensional CAD data is an example of a three-dimensional model (an image that can be recognized as a three-dimensional figure), and is not limited thereto.

  Note that the point data generation unit 112, the three-dimensional CAD data acquisition unit 115, the three-dimensional CAD data matching unit 116, and the three-dimensional CAD data generation unit 117 are not essential components (detailed later).

  Further, the three-dimensional CAD data generation unit 117 includes a point group data acquisition unit 111, a point data generation unit 112, a point group data matching unit 113, a point data deletion unit 114, a three-dimensional CAD data acquisition unit 115, Image processing different from the image processing apparatus provided with the control unit having the three-dimensional CAD data matching unit 116 (at least the point cloud data acquisition unit 111, the point cloud data matching unit 113, and the point data deletion unit 114) It may be provided in the apparatus. These two image processing apparatuses are connected by wire or wirelessly. Further, these two image processing apparatuses may be in the same or close place or in different places.

  FIG. 3 is a diagram illustrating a hardware configuration of the image processing apparatus 1. As shown in the figure, for example, an image processing apparatus 1 constituted by a computer or the like includes a CPU (Central Processing Unit) 101 that is an arithmetic device, a RAM (Random Access Memory) that is a volatile storage device, and a nonvolatile storage device. A memory 102 composed of a ROM (Read only Memory), an external storage device 103, a communication device 104 that communicates with an external device, and a user interface such as a keyboard, mouse, or touch panel that accepts input from the user An input device 105, an output device 106 such as a display, a read / write device 107 such as a CD drive and a DVD drive, and an interface (I / F) 108 for connecting the image processing apparatus 1 to other units.

  Each functional unit of the control unit 11 is realized, for example, when the CPU 101 reads a predetermined program stored in the memory 102 and executes the program. The predetermined program may be installed in advance in the memory 102, downloaded from a communication network (not shown) via the communication device 104, installed or updated, installed from the external storage device 103, or It may be updated, or may be installed or updated from a portable storage medium 109 such as a CD or DVD via the read / write device 107. The input unit 13 is realized by the input device 105. The output unit 12 and the communication unit 14 are realized by the communication device 104 and the I / F 108. The storage unit 15 is realized by the memory 102. The display device 30 is realized by the output device 106. The output device 106 is not essential as a configuration of the image processing apparatus 1 and may be connected via the I / F 108.

  The configuration of the image processing apparatus 1 described above is the main configuration in describing the features of the present embodiment, and is not limited to the above configuration.

  [Description of Operation] Next, the operation of the image processing apparatus 1 in this embodiment will be described.

  FIG. 4 is a flowchart showing the overall processing flow of the image processing apparatus 1 according to this embodiment. This process is started after a program execution start instruction is input by the input unit 13.

  The point cloud data acquisition unit 111 acquires first three-dimensional point cloud data 211 that is a result of measuring plant equipment from the first direction by the three-dimensional measuring device 20 (step S200). Further, the point cloud data acquisition unit 111 acquires second 3D point cloud data 212 that is a result of measuring the plant equipment from the second direction by the 3D measuring device 20 (step S202).

  FIG. 5 is an example showing how the three-dimensional measuring device 20 measures the switch 202 that is a plant facility. In the present embodiment, since the switch 202 is large, an example in which a part of the switch 202 surrounded by a one-dot chain line in FIG. In step S200, first three-dimensional point cloud data 211 that is a result of measuring the switch 202 from the direction A (first direction) is acquired. In step S202, second 3D point cloud data 212, which is a result of measuring the switch 202 from a direction B (second direction) that is different from the direction A, is acquired. Thereby, the result of having measured a part of switch 202 from two directions is acquirable.

  When the first three-dimensional point group data 211 and the second three-dimensional point group data 212 are acquired, the point data generation unit 112 displays the first three-dimensional point group data 211 and the second three-dimensional point group data. For each piece of point data included in 212, new point data is generated around the point data (step S204). The process of step S204 will be described in detail.

  FIG. 6 shows the position of the three-dimensional measuring device 20, an arbitrary point P included in the three-dimensional point group data 21 (first three-dimensional point group data 211, second three-dimensional point group data 212), and error. FIG. 6 is a diagram showing a relationship with a region R. The error region R is an error range between the position of the point P, which is a measured point, and the position of the object that is the basis of the measured point (point P). This error range is a three-dimensional region indicating a range of errors caused by the resolution of the laser beam used by the three-dimensional measuring device 20 to acquire the three-dimensional point cloud data 21. That is, the error region R can also be called an error range of the position of the point data. In the present embodiment, the three-dimensional point group data 21 formed by the three-dimensional measuring device 20 is acquired assuming that the point P, which is a measured point, is at the center of the error region R.

  As shown in FIG. 6, the error region R is mainly determined by the measurement resolution H in the plane direction and the measurement resolution V in the height direction. Here, the measurement resolution H in the plane direction and the measurement resolution V in the height direction are determined by the distance L between the measurement origin (here, the position of the three-dimensional measuring device 20) and the measurement object. For example, when L = 10 m, the measurement resolution H in the plane direction and the measurement resolution V in the height direction are about ± 2 mm (4 mm), but when L = 100 m, which is the use condition in the present embodiment. The measurement resolution H in the plane direction and the measurement resolution V in the height direction are about ± 20 mm (40 mm). When L = 120 m, which is the nominal measurement limit of the three-dimensional measuring instrument 20, the measurement resolution H in the plane direction and the measurement resolution V in the height direction are about ± 24 mm (48 mm). Note that the relationship between the distance L, the measurement resolution H in the plane direction, and the measurement resolution V in the height direction may be stored in advance in the storage unit 15 or the like, or may be acquired from the three-dimensional measuring instrument 20. Good. Thus, by determining the error region R based on the resolution, the error region R can be sized according to the distance L.

  The point data generation unit 112 acquires information regarding the distance L from the three-dimensional measuring device 20. Further, as shown in FIG. 6, the point data generation unit 112 sets an error region R around the point P based on the distance L, and generates a plurality of point data P1 within the range of the error region R. The point data generation unit 112 generates point data in the error region R so that the area close to the point P has a lot of point data P1 and the point data P1 decreases as the distance from the point P increases.

  FIG. 7 is a diagram illustrating a method in which the point data generation unit 112 generates a plurality of point data P1. FIG. 7A shows a density function of a standard normal distribution that uses the point data P1 for generation, and FIG. 7B shows a relationship between the density function of the standard normal distribution and the point data. The point data generation unit 112 generates a circle α and a circle β around the point P with σ and 3σ as radii in the density function of the standard normal distribution with the position of the point P as the origin. The circle α and the circle β are included in the error region R.

  Then, point data is generated at a high density inside the circle α, and point data is generated outside the circle α and inside the circle β at a lower density than the inside of the circle α. For example, the point data generation unit 112 generates ten pieces of point data each inside the circle α and outside the circle α and inside the circle β. Further, for example, the point data generation unit 112 generates point data with density D1 inside the circle α, and generates point data with density D2 (D1> D2) outside the circle α and inside the circle β. Good.

  In the present embodiment, the point data P1 is generated using the density function of the standard normal distribution, but the method of generating the point data P1 is not limited to this. As long as the point data generation unit 112 generates the point data P1 in the error region R such that the point data P1 increases in the area close to the point P and decreases as the distance from the point P increases, Such a method may be used.

  The error region R is not limited to this. The error region R may be a three-dimensional region determined based on the resolution of the laser beam used for acquiring the first point cloud data 21. For example, the measurement resolution H in the planar direction can be set larger than the exemplified size (for example, ± 20 mm when L = 100 m). In the example shown in FIG. 6, the plane parallel to the measurement resolution H in the plane direction of the error region is a curved surface, but it may be a plane.

  Returning to the description of FIG. The three-dimensional CAD data acquired by the three-dimensional CAD data acquisition unit 115 is the plant equipment data that is the basis of the first three-dimensional point cloud data 211 and the second three-dimensional point cloud data 212 acquired in steps S200 and S202. 22, the three-dimensional CAD data matching unit 116 generates the point data (step S204), and each of the first three-dimensional point group data 211 and the second three-dimensional point group data 212 is generated. On the other hand, matching (superposition) with the three-dimensional CAD data 22 is performed. The three-dimensional CAD data matching unit 116 constrains the point data that coincides with the surface of the three-dimensional CAD data 22 in the first three-dimensional point group data 211 and the second three-dimensional point group data 212 as a surface. (Surface constraint) (Step S206).

  Note that the plant equipment data that is the basis of the first three-dimensional point group data 211 and the second three-dimensional point group data 212 acquired in steps S200 and S202 is the three-dimensional data acquired by the three-dimensional CAD data acquisition unit 115. If it is not included in the CAD data 22, step S206 is not performed.

  Step S206 will be described with reference to FIGS. FIG. 8 is a diagram illustrating an H-shaped steel that is an example of a measurement target. The point cloud data acquisition unit 111 acquires the result of measuring the end face of the H-shaped steel as shown in FIG. 8 (measured from the H direction in FIG. 8) as the three-dimensional point cloud data 21. The three-dimensional CAD data 22 acquired by the three-dimensional CAD data acquisition unit 115 includes information on the shape of the end surface of the H-shaped steel, the dimensional ratio (here, x: y) of the end surface of the H-shaped steel, and the like.

  FIG. 9 is a diagram for explaining matching in the case where the end face of the H-shaped steel is measured. FIG. 9A shows three-dimensional point group data 21 obtained by measuring the end face of the H-shaped steel shown in FIG. FIG. 9B shows a state in which the three-dimensional CAD data 22 is matched with the three-dimensional point cloud data 21 shown in FIG. In order to make the figure easy to understand, in FIG. 9B, the three-dimensional point cloud data 21 is displayed as white circles or hatched circles. Further, the line in FIG. 9B indicates the three-dimensional CAD data 22.

  The three-dimensional point cloud data 21 shown in FIG. 9A is not H-shaped due to various measurement errors. On the other hand, the three-dimensional CAD data 22 is H-shaped because it is as in a catalog or the like. Since the three-dimensional CAD data 22 includes data indicating the shape of the end face of the H-shaped steel and the dimensional ratio (here, x: y) of the end face of the H-shaped steel, the three-dimensional CAD data matching unit 116 is a three-dimensional CAD data. Even if the size (absolute value) of the point cloud data 21 is not known, the 3D CAD data 22 can be matched with the 3D point cloud data 21.

  Note that the surface that the three-dimensional CAD data matching unit 116 constrains using the three-dimensional CAD data 22 is not limited to a flat surface, and may be a cylindrical surface or a curved surface.

  The three-dimensional CAD data matching unit 116 constrains the point data matched with the three-dimensional CAD data 22 in the three-dimensional point cloud data 21 as a surface so that the point cannot be moved or deleted. In FIG. 9B, the three-dimensional CAD data 22 has three surfaces h1, h2, and h3 that form an H shape. The three-dimensional CAD data matching unit 116 constrains the point data overlapping the surface h1 among the point data (hatched point data) overlapping the three-dimensional CAD data 22 as the surface h1, and the point data overlapping the surface h2 as the surface. Constrained as h2, and the point data overlapping the surface h3 is constrained as the surface h3.

  Further, consider the case where the side surface of the H-shaped steel as shown in FIG. Since the H-shaped steel is a metal, irregular reflection tends to occur particularly at the corners. Due to the irregular reflection, point data is measured at a position other than the originally measured position.

  FIG. 10 is a diagram for explaining matching when the side surface of the H-shaped steel is measured. FIG. 10A is a diagram for explaining that point data is measured at a position other than the originally measured position due to irregular reflection, and FIG. 10B is a three-dimensional measurement of the side surface of the H-shaped steel. A state in which the point cloud data 21 is matched with the three-dimensional CAD data 22 is shown.

  When a laser beam is irradiated from the measurement origin (here, the position of the three-dimensional measuring device 20) to an arbitrary point M at the corner of the H-shaped steel, the laser beam is diffusely reflected toward a point N not on the H-shaped steel. Since the three-dimensional measuring device 20 generates point data based on the laser beam reflected and returned from the object, in the case shown in FIG. The point data is generated at the position of the point N.

  As a result, as shown in FIG. 10B, the three-dimensional point group data 21 includes originally measured point data 21a and erroneous point data 21b due to irregular reflection or the like. In addition, when the wrong point data 21b is formed due to irregular reflection or the like, the point data is not measured at the position where it is originally measured.

  However, since the shape of the side surface of the H-shaped steel is known by matching the three-dimensional CAD data 22, the point data 21a to be measured and the erroneous point data 21b due to irregular reflection can be clearly distinguished. The point data matched with the three-dimensional CAD data 22 (for example, the point data 21a inside the three-dimensional CAD data 22) is constrained as a surface so that the points cannot be moved or deleted. Thereby, even when the point data 21a corresponding to the point data 21a is not included in the measurement result of the side surface of the H-shaped steel measured from the other direction (for example, the direction H) due to irregular reflection or the like, the point data 21a is deleted. The point data 21a can be used to generate three-dimensional data.

  Returning to the description of FIG. In the point cloud data matching unit 113, point data is generated by the point data generation unit 112 (step S204), and the surface is constrained by the three-dimensional CAD data matching unit 116 (step S206). The second three-dimensional point cloud data 212 is overlaid (step S208). Then, the point data deletion unit 114 deletes point data that do not match as a result of superimposing the first three-dimensional point group data 211 and the second three-dimensional point group data 212 (step S208) (step S210). Step S210 will be described with reference to FIG.

  FIG. 11 is a diagram for explaining point data deletion. FIG. 11A shows the first three-dimensional point group data 211, and FIG. 10B shows the second three-dimensional point group data 212.

  The first three-dimensional point cloud data 211 is data obtained by measuring the switch 202 from the direction A (side surface) shown in FIG. Therefore, irregular reflection occurs at the angle E of the switch 202, and the original point data (point data a) is included in the first three-dimensional point group data 211.

  The second three-dimensional point cloud data 212 is data obtained by measuring the switch 202 from the direction B (upper surface) shown in FIG. Therefore, the point data a is not included in the second three-dimensional point group data 212. However, due to the influence of reflection on the surface T, point data (point data c) that is not originally included is included in the second three-dimensional point group data 212.

  The point data deletion unit 114 deletes point data a, which is a point in the first three-dimensional point group data 211 and not in the second three-dimensional point group data 212, from the first three-dimensional point group data 211. Further, the point data deletion unit 114 deletes the point data c that is in the second 3D point cloud data 212 but not in the first 3D point cloud data 211 from the second 3D point cloud data 212. To do.

  The first three-dimensional point group data 211 and the second three-dimensional point group data 212 that are points that are in the first three-dimensional point group data 211 and not the second three-dimensional point group data 212. The point data c, which is a point not in 211, is determined by the point cloud data matching unit 113 superimposing the first 3D point cloud data 211 and the second 3D point cloud data 212 (step S208).

  Specifically, the point cloud data matching unit 113 determines the position of the object that is the basis of the point data for each point data of the first three-dimensional point cloud data 211. Further, the point cloud data matching unit 113 determines the position of the target object of the point data with respect to each point data of the second three-dimensional point cloud data 212. This is because each point data of the first three-dimensional point group data 211 and the measurement origin of the first three-dimensional point group data 211, and each point data of the second three-dimensional point group data 212 and the second three-dimensional This is possible by placing the measurement origin of the point cloud data 212 on the same coordinates.

  Then, the point group data matching unit 113 superimposes the first three-dimensional point group data 211 and the second three-dimensional point group data 212 so that any arbitrary one included in the first three-dimensional point group data 211 is included. Whether or not the point data obtained by measuring the position of the object that is the basis of the point data (first point data) is included in the second three-dimensional point group data 212, and included in the second three-dimensional point group data 212 It is determined whether the first three-dimensional point cloud data 211 includes point data obtained by measuring the position of an object that is a base of arbitrary point data (second point data). That is, the point cloud data matching unit 113 determines whether or not the point data obtained by measuring the same position of the object is included in the first 3D point cloud data 211 and the second 3D point cloud data 212. .

  It should be noted that the determination of whether the point data obtained by measuring the same position is included in the first three-dimensional point group data 211 and the second three-dimensional point group data 212 is not limited to the case of the completely same position, When the positions are substantially the same, such as being shifted by a minute error, the point data obtained by measuring the same position can be obtained.

  Then, the point data deleting unit 114 converts the point data obtained by measuring the position of the object serving as the basis of the first point data included in the first three-dimensional point group data 211 into the second three-dimensional point group data 212. If not included, the first point data is deleted from the first three-dimensional point cloud data 211. In addition, the point data deleting unit 114 converts the point data obtained by measuring the position of the object serving as the basis of the second point data included in the second three-dimensional point group data 212 into the first three-dimensional point group data 211. If not included, the second point data is deleted from the second three-dimensional point cloud data 212.

  FIG. 11C shows the result of deleting the point data a from the first three-dimensional point group data 211, and the point data c is deleted from the second three-dimensional point group data 212, that is, the first three-dimensional point group data 212. This is a result of removing noise from the dimensional point group data 211 and the second three-dimensional point group data 212. As a result, the first three-dimensional point group data 211 and the second three-dimensional point group data 212 include points included in both the first three-dimensional point group data 211 and the second three-dimensional point group data 212. Only data b is left.

  Thereby, noise can be removed from the three-dimensional point cloud data 21, and the accuracy of the three-dimensional point cloud data 21 can be increased.

  However, the point data deleting unit 114 is not included in the second 3D point cloud data 212 but in the first 3D point cloud data 211 or in the second 3D point cloud data 212. Even point data not included in the three-dimensional point cloud data 211 is deleted (not removed as noise) if the surface is constrained in step S206. Thereby, the accuracy of the three-dimensional point cloud data 21 can be further increased.

  Next, the three-dimensional CAD data generation unit 117 generates three-dimensional CAD data from the three-dimensional point cloud data 21 after removing noise (step S210), and displays the three-dimensional data generated by the output unit 12 on the display device 30. (Step S212). FIG. 13 is a display example on the display device 30. Thereby, highly accurate three-dimensional CAD data can be generated. Thus, the processing of the image processing apparatus 1 is finished.

  According to the present embodiment, in the 3D point cloud data measured from different directions, the point data that is not included in both of the 3D point cloud data measured from different directions is removed as noise. The accuracy of the point cloud data 21 can be increased.

  Further, in the present embodiment, by generating point data in the error region, the point data interval in the three-dimensional point group data, that is, the point data position error can be made smaller than the size of the error region. Therefore, the accuracy of the three-dimensional point cloud data 21 can be made higher than the measurement limit of the three-dimensional measuring device.

  In the present embodiment, the three-dimensional measuring device 20 is used to measure a large plant facility having a size of, for example, several tens to several hundreds of meters, for example, from a distance of about 100 m. Thus, by measuring from a distance of about 100 m, the error area (synonymous with the position data position error) becomes as large as about ± 20 mm. However, for example, in installation of a line or the like in the plant facility, only an error of about several mm (for example, 5 mm) is allowed. In this embodiment, since point data is generated (added) in the error region, the interval between the points can be reduced and the accuracy of the generated three-dimensional data can be improved.

  Further, in the present embodiment, point data that is not included in both of the three-dimensional point group data measured from different directions with respect to the data after the point data is generated in the error region is removed as noise. Therefore, the number of points included in both of the three-dimensional point group data can be increased, and the accuracy of the generated three-dimensional data can be improved.

  Moreover, in this Embodiment, since surface constraining is performed using 3D CAD data, the accuracy of the 3D point cloud data 21 can be further increased. In particular, when the surface is not constrained, point data that is not included in both of the 3D point cloud data is deleted, whereas the point data that is surface constrained is included only in one of the 3D point cloud data. Since it is not deleted even if it is not, it can be used as a basis for generating three-dimensional data. Therefore, it is possible to obtain three-dimensional point cloud data that can generate efficient and accurate three-dimensional CAD data.

  In the present embodiment, the point cloud data acquisition unit 111 is a three-dimensional point cloud data 21 (first three-dimensional point cloud data 211, second three-dimensional data) that is a result of measuring an object from two different directions. Although the point cloud data 212) is acquired, what the point cloud data acquisition unit 111 acquires is not limited to the three-dimensional point cloud data 21 that is a result of measuring an object from two different directions. For example, the point cloud data acquisition unit 111 may acquire 3D point cloud data 21 that is a result of measuring an object measured from three different directions.

  The point data generation unit 112 is not an essential configuration in the present invention. For example, in the present embodiment, the three-dimensional CAD data matching unit 116 matches the three-dimensional CAD data 22 with respect to the three-dimensional point group data 21 generated by the point data generation unit 112. The data that the three-dimensional CAD data matching unit 116 matches with the three-dimensional CAD data 22 may be the three-dimensional point cloud data 21 acquired by the point cloud data acquisition unit 111. Further, for example, in the present embodiment, the point cloud data matching unit 113 matches the plurality of 3D point cloud data 21 generated by the point data generation unit 112, but the point cloud data matching unit 113 The plurality of 3D point cloud data 21 to be matched may be the 3D point cloud data 21 acquired by the point cloud data acquisition unit 111. However, in order to increase the amount of information of points to be used and improve accuracy, the point cloud data matching unit 113 matches a plurality of three-dimensional point cloud data 21 after the point data generation unit 112 generates the point data. It is desirable to do.

  The three-dimensional CAD data acquisition unit 115 and the three-dimensional CAD data matching unit 116 are not essential components in the present invention. For example, in the present embodiment, the point cloud data matching unit 113 includes a plurality of points after the point data generation unit 112 generates point data and the 3D CAD data matching unit 116 performs matching with the 3D CAD data 22. Although the 3D point cloud data 21 is matched, the plurality of 3D point cloud data 21 matched by the point cloud data matching unit 113 may be the 3D point cloud data 21 acquired by the point cloud data acquisition unit 111. However, in order to increase the information amount of the points to be used and improve the accuracy, it is preferable that the point data deletion unit 114 deletes the point data from the three-dimensional point cloud data 21 after the surface constraint.

  Further, the three-dimensional CAD data generation unit 117 is not an essential configuration in the present invention. For example, if there is no three-dimensional CAD data generation unit 117, the output unit 12 outputs the three-dimensional point cloud data 21 from which the point data has been deleted by the point data deletion unit 114 to the display device 30 in step S212. Good. FIG. 13 is a display example when the display device 30 outputs the three-dimensional point cloud data 21. In this case, the display device 30 displays a collection of point data.

  In addition, this invention is not limited to an above-described Example, Various modifications are included. For example, the above-described embodiments have been described in detail for easy understanding of the present invention, and are not necessarily limited to those having all the configurations described (some configurations of the present embodiment are not limited). The image processing apparatus having the above is included in the present invention). It is also possible to add the configuration of another embodiment to the configuration of one embodiment. Further, it is possible to add, delete, and replace other configurations for a part of the configuration of each embodiment.

  Further, the present invention is not limited to the case where the image processing apparatus is provided. You may provide as a system which has a several image processing apparatus, and you may provide as a system which has an image processing apparatus and a three-dimensional measuring device. Moreover, you may provide as a three-dimensional measuring device containing an image processing apparatus.

  Further, the present invention is not limited to the case where it is applied to a plant such as a power plant described in the present embodiment, and can also be applied to construction sites such as elevator installation and repair, railway laying and repair, etc. . The present invention is particularly suitable for a large scale, measuring the entire site from a distance, and generating point cloud data that is the basis of a three-dimensional model, but is also applicable to small equipment such as NC machine tools. is there.

  In addition, each of the above-described configurations, functions, processing units, processing means, and the like may be realized in part or in whole by hardware. Each of the above-described configurations, functions, and the like may be realized by software by interpreting and executing a program that realizes each function by the processor. Information such as programs, tables, and files for realizing each function can be stored in a recording device such as a memory, a hard disk, or an SSD (Solid State Drive), or a recording medium such as an IC card, an SD card, or a DVD.

  Further, control lines and information lines indicate what is considered necessary for the explanation, and not all control lines and information battles are necessarily shown on the product. In practice, it may be considered that almost all the components are connected to each other.

1: Image processing device 11: Control unit 12: Output unit 13: Input unit 14: Communication unit 15: Storage unit 20: Three-dimensional measuring instrument 21: Three-dimensional point cloud data 22: Three-dimensional CAD data 30: Display device 101: CPU
102: Memory 103: External storage device 104: Communication device 105: Input device 106: Output device 107: Read / write device 108: I / F
111: Point cloud data acquisition unit 112: Point data generation unit 113: Point cloud data matching unit 114: Point data deletion unit 115: 3D CAD data acquisition unit 116: 3D CAD data matching unit 117: 3D CAD data generation unit 200: Power plant 201: Metal cladding 202: Switch 211: First three-dimensional point group data 212: Second three-dimensional point group data

Claims (10)

A point cloud data acquisition unit that acquires a measurement result of an object provided in a three-dimensional space as point cloud data that is an aggregate of point data, the first point cloud data acquired from a first direction; A point cloud data acquisition unit for acquiring second point cloud data acquired from a second direction different from the first direction;
When point data obtained by measuring the same position of the object is not included in the first point cloud data and the second point cloud data, a point cloud data deleting unit that deletes the point data;
An output unit for outputting the first point cloud data or the second point cloud data after being deleted by the point cloud data deletion unit;
An image processing apparatus comprising: The image processing apparatus according to claim 1,
For each piece of point data constituting the first point cloud data, a plurality of point data are generated within a predetermined area centered on the point data, and the second point cloud data is constituted For each of the data, a point data generation unit that generates a plurality of point data within a predetermined area centered on the point data,
The point cloud data deletion unit deletes the first point data from the first point cloud data generated by the point data generation unit, and the point data generation unit generates second point data. An image processing apparatus, wherein the second point data is deleted from point cloud data. The image processing apparatus according to claim 2,
The image processing apparatus, wherein the predetermined area is a three-dimensional area determined based on a resolution of a laser beam used for acquiring the first point group data or the second point group data. The image processing apparatus according to claim 1,
A three-dimensional data acquisition unit that acquires the three-dimensional data of the object and includes three-dimensional data including a shape of a surface of the object;
The point cloud data deletion unit superimposes the first point cloud data or the second point cloud data and the three-dimensional data, and the first point cloud data or the second point cloud data An image processing apparatus characterized by not deleting point data that coincides with a surface in the three-dimensional data. The image processing apparatus according to claim 1,
A three-dimensional model generation unit that generates a three-dimensional model based on the first point group data or the second point group data after being deleted by the point group data deletion unit;
The image processing apparatus, wherein the output unit outputs the generated three-dimensional model. A point cloud data acquisition unit that acquires a measurement result of an object provided in a three-dimensional space as point cloud data that is an aggregate of point data, the first point cloud data acquired from a first direction; A point cloud data acquisition unit for acquiring second point cloud data acquired from a second direction different from the first direction;
When point data obtained by measuring the same position of the object is not included in the first point cloud data and the second point cloud data, a point cloud data deleting unit that deletes the point data;
A first image processing apparatus comprising: an output unit that outputs the first point cloud data or the second point cloud data after being deleted by the point cloud data deletion unit;
A second image processing apparatus comprising a three-dimensional model generation unit that generates a three-dimensional model based on the first point cloud data or the second point cloud data output from the first image processing apparatus;
An image processing system comprising: A three-dimensional measuring instrument for measuring a measurement result of an object provided in the three-dimensional space as point cloud data that is an aggregate of point data;
An image processing system comprising an image processing device,
The image processing apparatus includes:
A point cloud data acquisition unit that acquires point cloud data measured by the three-dimensional measuring instrument, the first point cloud data acquired from the first direction, and a second direction different from the first direction A point cloud data acquisition unit for acquiring second point cloud data acquired from
When point data obtained by measuring the same position of the object is not included in the first point cloud data and the second point cloud data, a point cloud data deleting unit that deletes the point data;
An output unit for outputting the first point cloud data or the second point cloud data after being deleted by the point cloud data deletion unit;
An image processing system comprising: A measurement unit that measures the measurement result of the object provided in the three-dimensional space as point cloud data that is a collection of point data;
A point cloud data acquisition unit that acquires point cloud data measured by the three-dimensional measurement unit, the first point cloud data acquired from the first direction, and a second direction different from the first direction A point cloud data acquisition unit for acquiring second point cloud data acquired from
When point data obtained by measuring the same position of the object is not included in the first point cloud data and the second point cloud data, a point cloud data deleting unit that deletes the point data;
An output unit for outputting the first point cloud data or the second point cloud data after being deleted by the point cloud data deletion unit;
A three-dimensional measuring instrument comprising: A step of acquiring a measurement result of an object provided in a three-dimensional space as point cloud data that is an aggregate of point data, the first point cloud data acquired from the first direction; Obtaining second point cloud data obtained by acquiring the object from a second direction different from the first direction;
If point data obtained by measuring the same position of the object is not included in the first point cloud data and the second point cloud data, deleting the point data;
Outputting the first point cloud data after the first point data has been deleted or the second point cloud data from which the second point data has been deleted;
An image processing method comprising: A step of acquiring a measurement result of an object provided in a three-dimensional space as point cloud data that is an aggregate of point data, the first point cloud data acquired from the first direction; Obtaining second point cloud data obtained by acquiring the object from a second direction different from the first direction;
If point data obtained by measuring the same position of the object is not included in the first point cloud data and the second point cloud data, deleting the point data;
Outputting the first point cloud data after the first point data has been deleted or the second point cloud data from which the second point data has been deleted;
An image processing program for causing an arithmetic device to execute the above.

Images