JP2018014362A - Manufacturing device for electronic apparatus and cable shape estimation program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To implement accurate cable forming.SOLUTION: A manufacturing device for an electronic apparatus repeatedly executes processing, by a cable cross section/drawing direction calculation unit 26, of identifying a feature point while updating a drawing direction of a cable 44 on the basis of cross section shape data estimated from 3D point group data existing in a search volume and processing, by a search volume setting unit 24, of moving the search volume; and includes a cable shape estimation unit 30 for generating shape data (estimates a shape) on the cable 44 on the basis of a result of repeated processing. In this case, the cable cross section/drawing direction calculation unit calculates a bending radius r of the cable on the basis of the feature point; and when the bending radius r is shorter than a predetermined threshold (minimum bending radius R), does not update the latest main axis vector and skips the search volume by a predetermined distance in the latest main axis vector's direction via the search volume setting unit.SELECTED DRAWING: Figure 2

Description

  The present invention relates to an electronic device manufacturing apparatus and a cable shape estimation program.

  An electronic apparatus such as an information processing apparatus has a large number of components such as a hard disk device, a memory, a CPU, and a power supply device, and a housing for housing the components, and a wiring cable is provided in the housing.

  When manufacturing an electronic device, it is necessary to prevent the cable from getting in the way, for example, by turning the cable to a predetermined position (for example, near the wall of the housing). This operation is called cable forming. When cable forming is performed using a robot, it is necessary to accurately grasp the position and posture of the cable to be gripped by the robot.

  Conventionally, a technique for obtaining 3D data describing the shape of a flexible and long member and a technique related to the transfer of a workpiece by a robot are known (see, for example, Patent Documents 1 and 2). Moreover, the technique regarding the measurement of a three-dimensional shape is also known (for example, refer patent documents 3-5 etc.).

JP 2003-242185 A JP 2008-15683 A International Publication No. 2010/071139 Japanese Patent Laid-Open No. 2005-258643 JP 2007-80132 A

  When the shape of the cable is recognized from an image obtained by capturing the cable, since the cable is indefinite, it is difficult to perform a matching process using a straight line, an arc, or the like. For this reason, conventionally, a method of recognizing the shape of the cable by extracting and tracking an edge that is a boundary between the cable and the background from the end portion of the cable is employed.

  However, if another cable crosses the shape recognition target cable, or if the shape recognition target cable forms a ring and partly intersects, the cable at the intersection Cannot be tracked accurately, and the shape of the cable may be erroneously recognized. Further, when erroneous recognition occurs, there is a high possibility that cable forming using a robot will fail.

  In one aspect, the present invention provides an electronic device manufacturing apparatus capable of accurately performing an operation of shifting a cable to a predetermined state, and a cable shape estimation program capable of accurately estimating a cable shape. The purpose is to provide.

  In one aspect, the manufacturing apparatus of an electronic device has the cable corresponding to the lattice points in a two-dimensional plane from image data obtained by photographing the cable in a state where the cable is connected to a part of the component of the electronic device. An acquisition unit that acquires three-dimensional position data of a point on the surface; a setting unit that generates spatial data having a predetermined width in the extending direction of the cable in the vicinity of one end of the cable; and the existing in the spatial data Generates the cross-sectional shape data of the cable from the point where the three-dimensional position data is acquired, specifies the extending direction of the cable based on the cross-sectional shape data, and specifies the feature point satisfying a predetermined condition in the cross-sectional shape data The spatial data is moved based on the specified feature points and the extending direction, the process is repeatedly executed, and the shape data of the cable is determined from the processing result. A control unit that causes the gripping device to grip the cable based on a generation result of the processing unit, and moves the gripping device to move the cable. When the bending radius of the cable calculated based on the feature point of the cross-sectional shape data is less than a predetermined threshold, the feature point and the extension direction are discarded, and the spatial data is not discarded. An electronic device manufacturing apparatus that moves a predetermined distance along a stretching direction.

  As one aspect, it is possible to accurately perform the operation of transitioning the cable to a predetermined state.

It is the schematic which shows the structure of the manufacturing apparatus of the electronic device which concerns on one Embodiment. FIG. 2A is a diagram illustrating a hardware configuration of the cable recognition device, and FIG. 2B is a functional block diagram of the cable recognition device. It is a figure which shows 3D image and 3D point cloud data. FIG. 4A to FIG. 4C are views (No. 1) for explaining processing of the cable recognition device. It is FIG. (2) for demonstrating the process of a cable recognition apparatus. It is a flowchart which shows a series of processes performed by the manufacturing apparatus of an electronic device. Fig.7 (a)-FIG.7 (c) are figures which show the outline | summary of a series of processes of the manufacturing apparatus of an electronic device. It is a flowchart which shows the specific process of step S14 of FIG. It is a flowchart which shows the specific process of step S40 of FIG. 8, and S54. FIGS. 10A to 10D are diagrams (part 1) for explaining the processing of FIG. FIG. 11A to FIG. 11C are diagrams for explaining the processing of FIG. FIGS. 12A to 12C are diagrams (part 2) for explaining the processing of FIG. FIGS. 13A and 13B are diagrams (part 3) for explaining the processing of FIG. It is a figure which shows cable difference length D (i) typically (planar). FIG. 15A is a diagram illustrating an example of 3D point cloud data, and FIG. 15B is a diagram illustrating a part of FIG. It is a flowchart which shows the specific process of step S56 of FIG. It is a figure for demonstrating the bending radius r. FIG. 18A is a diagram illustrating a crossing state of cables, and FIG. 18B is a diagram illustrating a correct tracking direction and an incorrect tracking direction when the cables are crossing. FIG. 17 is a diagram (No. 1) for specifically describing the process of FIG. 16; FIGS. 20A to 20D are diagrams (part 2) for specifically explaining the processing of FIG. It is a figure for demonstrating a predetermined distance. It is a figure for demonstrating another example of a predetermined distance. FIG. 17 is a diagram (No. 1) for specifically describing a modified example of the process of FIG. 16; FIG. 24A to FIG. 24E are diagrams (No. 2) for specifically explaining a modified example of the processing of FIG.

  Hereinafter, an embodiment of an electronic device manufacturing apparatus will be described in detail with reference to FIGS.

  As shown in FIG. 1, the electronic device manufacturing apparatus 100 according to the present embodiment includes a robot 10 as a gripping device, a 3D camera 12, a cable recognition device 14, and a robot controller 16 as a control unit. ing.

  As an example, the robot 10 is a robot that performs cable forming while connecting the connectors 46A and 46B provided at one end and the other end of the cable 44 to the connector provided on the printed circuit board 42. The cable forming means a wiring operation in which the cable 44 is changed to a predetermined state. In the present embodiment, the cable forming is an operation in which the cable 44 is wound along a part of the outer periphery of the printed circuit board 42 as shown in FIG. The cable 44 is a flexible string-like member having a constant thickness and a circular cross section, and has an indefinite shape and is deformed three-dimensionally as needed. Note that there is continuity in the direction in which the cable 44 extends (stretching direction) (the bending has a certain curvature).

  The 3D camera 12 is, for example, a 3D point cloud camera, a 3D scanner, or the like, and captures a 3D image near the cable 44. The 3D image captured by the 3D camera 12 is transmitted to the cable recognition device 14.

  The cable recognition device 14 acquires 3D point cloud data from the 3D image captured by the 3D camera 12, recognizes the state of the cable 44 to be cable formed based on the 3D point cloud data, and outputs the recognition result to the robot controller 16. To do. In this case, the cable recognizing device 14 performs the tracking process of the cable 44 from one end (near the connector 46A) of the cable 44, and recognizes the posture and the existing position (XYZ coordinates) of a predetermined length from the one end.

  The robot controller 16 controls the robot 10 to perform an assembly (connection) operation of the cable 44 to the printed circuit board 42. Further, the robot controller 16 controls the robot 10 based on the recognition result of the cable recognition device 14 and executes cable forming for changing the cable 44 to a predetermined state (the state shown in FIG. 7C).

  Here, the cable recognition device 14 will be described in detail. FIG. 2A shows the hardware configuration of the cable recognition device 14. As shown in FIG. 2A, the cable recognition device 14 includes a CPU (Central Processing Unit) 90, a ROM (Read Only Memory) 92, a RAM (Random Access Memory) 94, and a storage unit (here, an HDD (Hard Disk Drive). )) 96, a network interface 97, a portable storage medium drive 99, and the like. Each component of the cable recognition device 14 is connected to a bus 98. In the cable recognition device 14, a program (including a cable shape estimation program) stored in the ROM 92 or the HDD 96, or a program (including a cable shape estimation program) read from the portable storage medium 91 by the portable storage medium drive 99. The function of each unit shown in FIG. 2B is realized by the CPU 90 executing.

  FIG. 2B shows a functional block diagram of the cable recognition device 14. When the CPU 90 executes the program, the cable recognition device 14 has a cable 3D data acquisition unit 20 as an acquisition unit, a cable end extension direction measurement unit 22 as a specification unit, and a setting unit as illustrated in FIG. It functions as a search volume setting unit 24, a cable cross-section / extension direction calculation unit 26, a point-of-interest update unit 28, a cable shape estimation unit 30, and a communication unit 32.

  The cable 3D data acquisition unit 20 acquires image data (3D image) taken by the 3D camera 12. Note that the 3D image is an image of at least the entire cable 44 to be cable formed (tracked). Further, the cable 3D data acquisition unit 20 acquires three-dimensional position data (referred to as “3D point cloud data”) of points on the surface of the cable 44 corresponding to lattice points in the two-dimensional plane from the acquired 3D image. To do. FIG. 3 shows an example of a 3D image. The lattice points shown in FIG. 3 are points corresponding to the pixels of the 3D camera 12. As shown in FIG. 3, the cable 3D data acquisition unit 20 acquires the height position (Z coordinate) of the surface of the cable 44 for each lattice point in the two-dimensional plane (XY plane) corresponding to the pixel of the 3D camera 12. The three-dimensional data including the XY coordinates of each lattice point and the acquired Z coordinates are set as 3D point group data.

  FIG. 15A shows an example of 3D point cloud data (an image obtained by enlarging the vicinity of the cable 44), and FIG. 15B shows a part of FIG. In other words, a state where the cross section of the cable 44 is viewed vertically is shown. As indicated by a dotted line in FIG. 15B, it is apparent that the 3D point group data includes data of the upper half (arc-shaped portion) of the cross section of the cable 44.

  Returning to FIG. 2B, the cable end point extension direction measurement unit 22 sets a point in the vicinity of one end of the cable 44 as a point of interest from the 3D image data, and specifies a direction (extension direction) in which the cable 44 extends at the point of interest. The “target point” is a reference point for searching the shape of the cable 44. The cable end extension direction measuring unit 22 specifies the position and orientation of the connector 46A from the 3D image by model matching using a 3D CAD model of the connector 46A prepared in advance. The cable end extension direction measuring unit 22 determines a predetermined position of the cable determined from the position and posture of the connector 46A (for example, the highest position of the boundary portion between the cable 44 and the connector 46A as shown in FIG. 4A). Let it be a point of interest. Further, the cable end point extension direction measuring unit 22 specifies the extension direction of the cable 44 at the point of interest based on the posture of the connector 46A. Here, a vector having a predetermined length defined by the point of interest and the extending direction is referred to as a principal axis vector (see FIG. 4A).

  The search volume setting unit 24 generates (sets) a search volume as a range to be noted in order to estimate the cross-sectional shape of the cable 44. The search volume is rectangular parallelepiped space data including a point of interest inside and having a predetermined width in the extending direction of the cable 44. Specifically, the search volume is a space having a predetermined width a having a square surface with one side b as shown by a dotted line in FIG. Here, the predetermined width a of the search volume (see FIG. 4B) is, for example, about 1.5 times the size of one pixel of the camera (the distance between lattice points in FIG. 3), and one side of the square surface. The dimension b is assumed to be a dimension obtained by adding a predetermined value α to the diameter of the cable 44.

  The cable cross section / extension direction calculation unit 26 generates cross section data of the cable 44 from the 3D point cloud data existing in the search volume, and newly calculates the extension direction of the cable 44 based on the cross section data ( (Refer FIG.4 (c)). Further, the characteristic point of the cable 44 (the point at the highest position among the 3D point cloud data existing in the search volume) is specified. The cable cross-section / extension direction calculation unit 26 stores the calculated extension direction and information on the identified feature points (information on the principal axis vector) in the data storage unit 36. Further, the cable cross-section / extension direction calculation unit 26 determines whether another cable or the tracked cable 44 itself crosses the cable forming target cable 44 (tracked cable). When it is determined that they intersect, the cable cross section / stretch direction calculation unit 26 cooperates with the search volume setting unit 24 to execute processing so as not to be affected by the intersection.

  The point-of-interest update unit 28 sets a new point of interest as shown in FIG. 4C based on the new extension direction and feature points calculated by the cable cross-section / extension direction calculation unit 26 (the point of interest is changed). Update).

  The cable shape estimation unit 30 generates (estimates) the shape data of the cable 44 based on the results of the search volume setting unit 24, the cable cross-section / extension direction calculation unit 26, and the point-of-interest update unit 28 repeatedly executing the processing. That is, as shown in FIG. 5, the cable shape estimation unit 30 obtains information on the principal axis vector of the cable 44, which is sequentially calculated using the search volume sequentially set along the extending direction of the cable 44, from the data storage unit 36. get. And the cable shape estimation part 30 pinpoints the location away from the predetermined length m (refer Fig.7 (a)) from the edge part of the cable 44 based on the acquired information. Furthermore, the cable shape estimation unit 30 estimates the position (XYZ coordinates) where the specified location exists and the attitude of the cable 44 at the location, that is, the position and orientation of the cable 44 that the robot 10 should grip in cable forming.

  The communication unit 32 transmits an instruction based on the estimation result of the cable shape estimation unit 30 to the robot controller 16. The robot controller 16 controls the robot 10 to cause the robot 10 to grip the cable 44 and move the robot 10 that has gripped the cable 44, thereby causing the cable 44 to transition to a predetermined state (executes cable forming). .

(Processing of electronic device manufacturing apparatus 100)
Next, processing executed by the electronic apparatus manufacturing apparatus 100 according to the present embodiment will be described in detail.

  FIG. 6 is a flowchart showing a series of processes executed by the electronic device manufacturing apparatus 100. Hereinafter, an outline of processing of the electronic device manufacturing apparatus 100 will be described with reference to FIG.

  In the process of FIG. 6, first, in step S <b> 10, the robot controller 16 drives the robot 10 according to a predetermined procedure, and inserts the connectors 46 </ b> A and 46 </ b> B at both ends of the cable 44 into the connectors on the printed circuit board 42. As a result, the cable 44 is connected to the printed circuit board 42. The operation of connecting the cable 44 to the printed circuit board 42 may be performed by a robot different from the robot that performs cable forming, or may be performed by a person (operator). FIG. 7A shows a state where the cable 44 is connected to the printed circuit board 42.

  Next, in step S14, the cable recognizing device 14 executes cable recognition processing. The cable recognition process is executed according to the flowchart of FIG. 8, and details will be described later.

  Next, in step S16, the cable recognizing device 14 determines the position and orientation information (specifically, the vector V (n) shown in FIG. 7A) at a location having a predetermined length (m) from one end of the cable 44. Information). The cable recognition device 14 transmits the acquired information on the position and orientation of the cable 44 to the robot controller 16.

  Next, in step S <b> 18, the robot controller 16 moves the hand of the robot 10 from one end of the cable 44 to a predetermined length (m) based on the information received from the cable recognition device 14. Next, in step S <b> 20, the robot controller 16 holds the cable 44 with the hand of the robot 10. FIG. 7B shows a state where the hand of the robot 10 holds the cable 44. The robot controller 16 controls the posture of the robot 10 so that the hand of the robot 10 opens in a direction substantially perpendicular to the vector V (n) when gripping the cable 44.

  Next, in step S <b> 22, the robot controller 16 drives the robot 10 to the position shown in FIG. 7C and hooks the cable 44 on the corner of the printed circuit board 42.

  Next, in step S <b> 24, the robot controller 16 controls the robot 10 to release the grip of the cable 44 and retract the robot 10 from the printed circuit board 42.

  Thus, a series of processes by the electronic apparatus manufacturing apparatus 100 is completed.

(About processing of step S14)
Next, the process of step S14 (cable recognition process) will be described in detail along the flowcharts of FIGS. 8, 9, and 16 with reference to other drawings as appropriate.

  In the process of FIG. 8, first, in step S30, the cable end extension direction measuring unit 22 acquires a 3D image from the 3D camera 12, and the connector 46A provided at one end of a cable (cable to be tracked) whose shape is estimated. The 3D posture is measured. In this case, the position and orientation of the connector 46A are specified by model matching using a 3D CAD model of the connector 46A prepared in advance.

  Next, in step S <b> 32, the cable end point extension direction measuring unit 22 acquires the principal axis vector of the cable end. Specifically, the cable end extension direction measuring unit 22 takes the highest position of the boundary between the cable 44 and the connector 46A as shown in FIG. The extension direction of the cable 44 is specified from the information, and a vector defined by the point of interest and the extension direction is set as the principal axis vector of the cable end. The cable end extension direction measuring unit 22 can specify, for example, the direction in which a specific side of the connector 46A extends as the extension direction of the cable 44.

  Next, in step S <b> 34, the cable 3D data acquisition unit 20 acquires 3D point cloud data from the 3D camera 12.

  Next, in step S36, the search volume setting unit 24 sets a search volume using the principal axis vector at the cable end. As shown in FIG. 10A, the search volume setting unit 24 sets a rectangular parallelepiped search volume including a cable end (point of interest) and having a predetermined width a in the direction of the principal axis vector. In FIG. 10A, white circles indicate 3D point cloud data.

  Next, in step S38, the cable cross section / extension direction calculation unit 26 sets the counter i to zero.

  Next, in step S40, the cable cross-section / extension direction calculation unit 26 executes a cross-section search process. In step S40, processing according to the flowchart of FIG. 9 is executed.

  In the process of FIG. 9, the cable cross section / extension direction calculation unit 26 first extracts points in the search volume from the 3D point cloud data in step S70. In the example of FIG. 10A, it is assumed that 3D point cloud data indicated by black circles included in the search volume is extracted as shown in FIG. 10B.

  Next, in step S72, the cable cross-section / stretch direction calculation unit 26 performs elliptic fitting using the extracted points, as indicated by the one-dot chain line in FIG. 10C, and obtains the cross-sectional shape data (elliptical shape data) of the cable. obtain. FIG. 11A shows an example of a 3D image of the cable 44. For example, it is assumed that the points extracted in step S70 are indicated by black circles. In this case, since the cross section including the extracted point (the cross section taken along the line AA and the YZ cross section) is deviated from the vertical cross section (the cross section taken along the line BB) of the cable 44, as shown in FIG. It becomes oval. On the other hand, the vertical cross section becomes a perfect circle as shown in FIG.

  Next, in step S74, the cable cross-section / extension direction calculation unit 26 extracts a point at the highest position among the extracted points as a feature point P (i) (here, P (0)). In FIG. 10D, the feature point P (0) is indicated by a black circle.

  Next, in step S76, the cable cross section / extension direction calculation unit 26 calculates the vertical cross section of the cable 44 from the obtained distortion of the elliptical data. Next, in step S78, the cable cross section / extension direction calculation unit 26 sets the normal vector of the vertical cross section of the cable 44 as the principal axis vector V (i) (here, V (0)). In these steps S76 and S78, the cable cross-section / stretch direction calculation unit 26 calculates the elliptical plane normal vector from the major axis, minor axis ratio and major axis and minor axis directions of the elliptical shape data acquired in step S72. The difference (angle θ in FIG. 11A) from the vertical cross section (round surface) is calculated. Then, the cable cross section / extension direction calculation unit 26 sets a vector obtained by shifting the vector extending in the lattice direction (X direction in FIG. 11A) by the difference (θ) as the main axis vector V (i). The cable cross-section / extension direction calculation unit 26 stores the feature point extracted in step S74 and the spindle vector information extracted in step S78 in the data storage unit 36. After step S78, the process proceeds to step S46 in FIG.

  In step S46, the point-of-interest update unit 28 increments i by 1 (i = i + 1).

  Next, in step S48, the point-of-interest update unit 28 sets a small search volume centered on the principal axis vector V (i-1) (here, V (0)). As shown in FIG. 12A, the small search volume is assumed to be a rectangular parallelepiped space having a square section (or rectangle) with the main axis vector V (0) extending from the feature point P (0) as the central axis. . The small search volume may be cylindrical space data, for example.

  Next, in step S50, the target point update unit 28 searches for a point (new target point) included in the small search volume. Here, it is assumed that a point indicated by a black circle in FIG. 12B is searched as a new point of interest. The size of the small search volume is preferably set so as not to include a plurality of points. However, the present invention is not limited thereto, and when a plurality of points are included in the small search volume, for example, a point closest to the feature point P (i) among the plurality of points may be searched for as a new point of interest. .

  Next, in step S52, the search volume setting unit 24 sets a search volume including a searched point of interest and having a predetermined width in the direction of the principal axis vector V (i−1) (here, V (0)). . FIG. 12B shows a newly set search volume.

  Next, in step S54, the cable cross-section / extension direction calculation unit 26 executes a cross-section search process (FIG. 9). Note that the cross-section search process is executed in the same procedure as described above. Therefore, the cross-section search process extracts points in the search volume as indicated by black circles in FIG. 12C, and obtains feature points P (i) and principal axis vectors V (i). Note that the point of interest searched for in step S50 (black dot in FIG. 12B) and the feature point P (i) may be the same or different. FIG. 13A shows a state in which feature points P (i) (= P (1)) indicated by black circles are acquired based on the elliptical shape data, and FIG. A state in which the principal axis vector V (i) (= V (1)) is acquired based on the elliptical shape data is shown.

  Next, in step S55 of FIG. 8, the cable cross-section / extension direction calculation unit 26 determines whether two feature points are obtained before the current feature point (P (i)). If the determination in step S55 is negative, that is, if i <2, the cable bending radius that needs to be calculated in step S56 described later cannot be calculated, and the presence or absence of cable crossing cannot be determined. The cable cross section / extension direction calculation unit 26 skips step S56 and proceeds to step S57.

In step S57, the cable shape estimation unit 30 calculates the cable difference length D (i) between the previous feature point P (i-1) and the current feature point P (i). In this case, the XYZ coordinates of the feature point P (i-1) are (x (i-1), y (i-1), z (i-1)), and the XYZ coordinates of the feature point P (i) are (x (i), y (i), z (i)), D (i) can be calculated using the following equation (1).
D (i) = {(x (i−1) −x (i)) ² + (y (i−1) −y (i)) ² + (z (i−1) −z (i)) ² } ^1/2 (1)

  In FIG. 14, the cable differential length D (i) is schematically (planarly) shown.

  Next, in step S58, the cable shape estimation unit 30 calculates the sum of the cable differential lengths from the cable end (feature point P (0)) to the i-th feature point (P (i)), and L (i) To do. That is, the cable shape estimation unit 30 calculates the sum of D (1), D (2),..., D (i) and sets it as L (i).

  Next, in step S60, the cable shape estimation unit 30 determines whether or not the value of the length L (i) is larger than the value of the length m. Here, the length m is assumed to be the length of one side of the printed circuit board 42 (the short side in FIG. 7A), as shown in FIG. 7A. That is, in step S60, as shown in FIG. 7C, it can be said that it is determined whether or not the shape of the cable 44 has been traced to a position located at the corner of the printed circuit board 42 during cable forming. If the determination in step S60 is negative, the cable shape estimation unit 30 returns to step S46 and repeatedly executes the processes and determinations in steps S46 to S60.

(About processing of step S56)
Next, the process of step S56 executed when the determination of step S55 is affirmed will be described in detail along the flowchart of FIG. In step S56, the cable cross section / extension direction calculation unit 26 and the search volume setting unit 24 execute a cable crossing recognition process. In this process, it is determined whether or not the cables intersect based on the result of calculating the bend radius of the cable being tracked. If the cables intersect, the process considering the intersection is executed.

  In the process of FIG. 16, first, in step S102, the cable cross-section / extension direction calculation unit 26 calculates the current feature point P (i) and the previous two feature points P (i-1) and P (i-2). The bending radius r is calculated from the three points. In this case, as shown in FIG. 17, a circle equation is obtained from the coordinates of the three feature points P (i) to P (i-2), and the radius of the circle indicated by the obtained equation is set as a bending radius r.

  Next, in step S104, the cable cross-section / extension direction calculation unit 26 determines whether or not the bending radius r is equal to or larger than the minimum bending radius R of the cable (r ≧ R). The case where the determination in step S104 is affirmed means that the possibility that the cables cross each other is low. In this case, the process in FIG. 16 is terminated, and the process proceeds to step S57 in FIG.

  On the other hand, if the determination in step S104 is negative, that is, if the bending radius r is less than the minimum bending radius R, the cable cross-section / extension direction calculation unit 26 proceeds to step S106. The case of shifting to step S106 means that there is a high possibility that the cable intersects the cable being tracked (the cable for which the shape is estimated) as shown in FIG. Yes. If the cables cross, the cable should be tracked along the correct tracking direction indicated by the solid arrow in FIG. 18B (the search volume should be moved), but along the wrong tracking direction (broken arrow direction). May track the cable. Therefore, in the present embodiment, when the cables cross each other, the processes of steps S106 to S110 in FIG. 16 are executed in order to correctly track the cables.

  In step S106, the cable cross-section / extension direction calculation unit 26 discards the latest spindle vector (not stored in the data storage unit 36), and sets the previous spindle vector as the latest spindle vector. That is, the cable cross-section / extension direction calculation unit 26 does not update the latest spindle vector with the spindle vector V (i) obtained immediately before.

  Next, in step S108, the cable cross-section / extension direction calculation unit 26 determines whether or not the determination in step S104 is continuously denied. If the determination in step S108 is negative, the process proceeds to step S110. If the determination is positive, the process in FIG. 16 is terminated, and the process proceeds to step S57 in FIG.

  If the determination in step S108 is negative and the process proceeds to step S110, the search volume setting unit 24 skips the search volume by a predetermined distance in the direction of the latest spindle vector under the instruction of the cable cross section / extension direction calculation unit 26. A specific example of the predetermined distance will be described later.

  Here, the processing of FIG. 16 will be described in detail with reference to FIGS. 19 and 20 show a state in which the extension direction of the cable being tracked coincides with the Y-axis direction for simplification of description.

  For example, in FIG. 19, when there is a search volume at the position indicated by (1) (S52), all of the 3D point cloud data in the search volume is point data on the cable being tracked. Therefore, in this case, elliptical shape data reflecting the shape of the cable being tracked as shown by a broken line in FIG. 20A is obtained (S72). As shown in FIGS. 19 and 20A, the main axis vector in this case is a main axis vector V (s) having the feature point P (s) as a base point (S74 to S78). The cable cross-section / stretch direction calculation unit 26 stores the acquired spindle vector data (including feature point data) in the data storage unit 36.

  Next, when the search volume is moved to the position indicated by (2) in FIG. 19 (S52), the 3D point cloud data in the search volume intersects with the data (●) of the point on the cable being tracked. The data (△) of the points on the cable is included. In this case, when the 3D point cloud data is elliptically fitted, elliptical shape data that does not reflect the shape of the cable being tracked is obtained as shown by a broken line in FIG. 20B (S72), and the feature point P (s + 1) is obtained. ) As a base point is obtained (S74 to S78). However, in this case, the direction of the principal axis vector V (s + 1) is extremely different from V (s), and the bending radius r is less than the minimum bending radius R (S104: negative). (S + 1) and the feature point P (s + 1) are discarded (not stored in the data storage unit 36), and V (s) remains as the latest spindle vector (S106). Further, the search volume is skipped by a predetermined distance along the direction of the principal axis vector V (s) and moved to the position indicated by (3) in FIG. 19 (S110).

  Next, when the search volume is moved to the position indicated by (3) in FIG. 19, the 3D point cloud data in the search volume includes the data of the point on the cable being traced (●) and the intersecting cable. Data for the upper point (Δ) is included. In this case, when the 3D point group data is elliptically fitted, elliptical shape data that does not reflect the shape of the cable being tracked is obtained, as indicated by the broken line in FIG. 20C, and the feature point P (s + 2) is used as the base point. A principal axis vector V (s + 2) is obtained. However, in this case, the direction of the principal axis vector V (s + 2) is extremely different from that of V (s), and the bending radius r is less than the minimum bending radius R (S104: negative). (S + 2) and the feature point P (s + 2) are discarded (not stored in the data storage unit 36), and V (s) remains as the latest spindle vector (S106). At this stage, since the determination in step S104 is continuously denied, the search volume does not skip a predetermined distance and proceeds to step S57 in FIG.

  Next, when the search volume is moved to the position indicated by (4) in FIG. 19 (S52), all of the 3D point cloud data in the search volume is the data of the points on the cable being tracked. Ellipse shape data reflecting the shape of the cable being tracked as shown by the broken line in (d) is obtained (S72). In this case, a principal axis vector V (s + 3) having the feature point P (s + 3) as a base point is obtained. Here, the direction of the principal axis vector V (s + 3) approximates to V (s), and the bending radius r becomes equal to or larger than the minimum bending radius R (S104: affirmative), so that the latest principal axis vector is the principal axis vector V (s + 3). (The data is stored in the data storage unit 36).

  If the search volume is skipped as shown in FIG. 19, the cable shape estimation unit 30 immediately before being stored in the feature point P (s + 3) and the data storage unit 36 in step S57 of FIG. The difference length from the feature point P (s) is calculated.

  Here, an example of the distance (predetermined distance) at which the search volume is skipped in step S110 will be described with reference to FIGS.

  When the cables cross as shown in FIG. 21, the overlapping width in the Y-axis direction is minimized when the cables are orthogonal. Further, the width at that time is equal to the lateral width Q (the diameter if the cross section of the cable is circular) Q of the intersecting cables (the cables located on the upper side). Therefore, the cable cross section / extension direction calculation unit 26 sets Q as a predetermined distance for skipping the search volume. As a result, it is possible to reduce the amount of processing (the number of times of processing using the search volume) while maintaining the cable shape estimation accuracy. The intersecting cables may be other cables located in the vicinity, or may be the cables being tracked themselves. When there are a plurality of cables that may cross like this, the search volume may be skipped by the smallest value among the diameters of the plurality of cables.

  As a method for skipping the search volume, a method as shown in FIG. 22 may be employed. That is, the cable cross section / extension direction calculation unit 26 first sets an edge extraction region extending in the direction of the latest principal axis vector in the vicinity of the current search volume, as shown in FIG. Then, the cable cross section / extension direction calculation unit 26 extracts the edge candidates of the intersecting cables in the edge extraction region of the image data. Here, the edge candidate means a portion estimated as a boundary portion between the cable and the background, and can be estimated based on a change in lightness or a color of image data. Next, the cable cross-section / extension direction calculation unit 26 extracts, as two edges, parallel edge candidates that are separated from each other by the lateral width (diameter) Q of the intersecting cables. In FIG. 22B, the two extracted edges are indicated by bold lines. Next, the cable cross-section / extension direction calculation unit 26 skips the search volume through the search volume setting unit 24 to the point where the latest principal axis vector and the edge far from the current search volume intersect. By skipping the search volume in this way, it is possible to reduce the amount of processing (the number of times of processing using the search volume) while maintaining the accuracy of cable shape estimation as in the case of FIG.

  Returning to FIG. 8, when the processes and determinations in steps S46 to S60 are repeatedly performed and the determination in step S60 is affirmative, that is, the cable 44 can be traced from one end of the cable 44 to a position farther than the length m. In the case, the process proceeds to step S62. Here, as an example, it is assumed that the value of i at the stage where L (i) exceeds m is n. In step S <b> 62, the communication unit 32 transmits the spindle vector V (i) (here, V (n)) to the robot controller 16 under the instruction of the cable shape estimation unit 30. Note that the data of the principal axis vector V (n) includes the coordinate value of the feature point P (n).

  Thus, the cable recognition process in step S14 is completed.

  Note that the robot controller 16 that has acquired the data of the spindle vector V (n) executes the processing after step S16 in FIG. As a result, the robot 10 can accurately perform processing (cable forming) for causing the cable 44 to transition to the state of FIG. 7C based on the accurately estimated position and orientation of the cable.

  In the above embodiment, the search volume setting unit 24, the cable cross-section / stretching direction calculation unit 26, the point-of-interest update unit 28, and the cable shape estimation unit 30 function as a processing unit that generates cable shape data. It has been realized.

  As described above in detail, according to the present embodiment, the cable 3D data acquisition unit 20 acquires 3D point cloud data of points on the surface of the cable 44 from the 3D image obtained by capturing the cable 44, and the search volume. The setting unit 24 sets (generates) a search volume in the vicinity of one end of the cable 44 from the 3D image. Further, the cable shape estimation unit 30 is based on the result of repetition of the process of the cable cross section / extension direction calculation unit 26 (main axis vector update process) and the process of the search volume setting unit 24 (process of moving the search volume). The shape data of the cable 44 is generated (estimated shape). In this case, the cable cross-section / stretching direction calculation unit 26 calculates the cable bending radius r based on the feature points, and if the bending radius r is less than a predetermined threshold (minimum bending radius R), the latest spindle. The search volume is skipped by a predetermined distance in the direction of the latest spindle vector via the search volume setting unit 24 without updating the vector. In this way, when the bending radius r of the cable is less than the minimum bending radius R, the latest principal axis vector is not updated. Therefore, even when the cables intersect, the tracked cable is accurately tracked and the shape Can be estimated. Further, by skipping the search volume, it is possible to reduce the amount of processing when tracking the cable (the number of elliptic fittings using the search volume). Further, in the present embodiment, by performing elliptical fitting using 3D point cloud data, the shape of the cable 44 can be traced from the vicinity of one end along the extending direction without using the edge portion of the cable 44. Therefore, the shape of the cable 44 can be accurately identified without being affected by noise or the like. Furthermore, cable forming can be performed accurately by using the shape specified with high accuracy.

  In this embodiment, the distance for skipping the search volume is the same distance as the diameter Q of the intersecting cables. As a result, the processing amount can be reduced without reducing the accuracy of cable shape estimation.

  In the present embodiment, a connector 46A is provided at one end of the cable 44, and the cable end extension direction measuring unit 22 is based on the position and orientation of the connector 46A specified from the 3D image, and a point of interest near one end of the cable. And the stretching direction. In this way, by using the position and orientation of the connector 46A, which is a fixed object, the point of interest near the one end of the irregular cable 44 and the extending direction can be accurately identified.

  In the present embodiment, the cable shape estimation unit 30 estimates the posture and position of a portion exceeding a predetermined length (m) from one end of the cable 44. As a result, information necessary for cable forming can be estimated.

  In the present embodiment, the point-of-interest update unit 28 uses a small search volume as shown in FIG. 12A when updating the point of interest. As a result, a new point of interest can be set with a simple process compared to a case where a point through which the main axis vector passes or a point having the shortest distance from the main axis vector is used as a new point of interest.

  In the above-described embodiment, the case where the 3D camera 12 is used has been described. However, the present invention is not limited thereto, and a camera capable of capturing a two-dimensional image may be used. In this case, the 3D point group data may be generated by detecting the height position of the point on the cable surface corresponding to the lattice point based on the color density or luminance of the two-dimensional image.

  In the above embodiment, the case where the connectors 46A and 46B are provided at both ends of the cable 44 has been described. However, the present invention is not limited to this. That is, both ends of the cable 44 may be soldered to the printed circuit board 42 or the like. In this case, when the end portion of the cable 44 is connected substantially perpendicular to the surface of the printed circuit board 42, the cable end point extension direction measuring unit 22 extends the feature point at the cable end in step S32 of FIG. The direction may be set to the normal direction of the printed circuit board 42.

  In the above embodiment, as shown in FIG. 20, the case where the elliptical fitting is performed using all of the 3D point cloud data in the search volume has been described, but the present invention is not limited to this. For example, ellipse fitting may be performed by a method as shown in FIGS.

  FIG. 23 shows search volumes (1) to (5), and FIGS. 24 (a) to (e) show examples of elliptic fittings corresponding to search volumes (1) to (5). Yes.

  In FIG. 23, when there is a search volume at the position indicated by (1), the 3D point cloud data in the search volume is data of the point on the cable being tracked. Such elliptical shape data is obtained. The principal axis vector in this case is indicated by V (s) in FIG. Next, when the search volume is moved to the position indicated by (2) in FIG. 23, the 3D point cloud data in the search volume includes the data of the point on the cable being traced (●) and the intersecting cable. Data for the upper point (Δ) is included. In this case, by performing noise removal processing or the like at the time of ellipse fitting, as shown in FIG. 24B, the elliptical fitting of the tracked cable cross section and the elliptical fitting of the crossing cable section are performed. Can do. In FIG. 24 (b), the principal axis vector V (s + 1) of the ellipse having the characteristic point that increases the cable bending radius among the feature points of the two ellipses is set as the latest principal axis vector, and the other principal axis vector V is obtained. (K + 1) is discarded.

  Next, when the search volume is moved to the position indicated by (3) in FIG. 23, the 3D point group data in the search volume includes only the data (Δ) of the points on the intersecting cables. In this case, when the 3D point group data is elliptically fitted, elliptical shape data as indicated by a broken line in FIG. 24C is obtained, and a principal axis vector V (k + 2) is obtained. However, in this case, the direction of the principal axis vector V (k + 2) is extremely different from the latest principal axis vector V (s + 1), and the bending radius r is less than the minimum bending radius R (S104: No). The spindle vector V (k + 2) is discarded, and V (s + 1) is set as the latest spindle vector (S106). Further, the search volume is skipped by a predetermined distance along the direction of the principal axis vector V (s + 1) and moved to the position indicated by (4) in FIG. 23 (S110).

  Next, when the search volume is moved to the position indicated by (4) in FIG. 23, the 3D point cloud data in the search volume includes the data of the point on the cable being traced (●) and the intersecting cable. Data for the upper point (Δ) is included. In this case, by performing noise removal processing or the like at the time of elliptical fitting, as shown in FIG. 24 (d), the elliptical fitting of the tracked cable cross section and the elliptical fitting of the crossing cable section are performed. Can do. In FIG. 24 (d), the principal axis vector V (s + 3) of the ellipse having the characteristic point that increases the cable bending radius among the feature points of the two ellipses is set as the latest principal axis vector, and the other principal axis vector V is obtained. Discard (k + 3).

  Next, when the search volume is moved to the position indicated by (5) in FIG. 23, since the 3D point cloud data in the search volume is data of the point on the cable being tracked, the broken line in FIG. Ellipse shape data as shown in FIG. The principal axis vector in this case is indicated by V (s + 4) in FIG. In this case, the direction of the principal axis vector V (s + 4) approximates to V (s + 3), and the bending radius r becomes equal to or greater than the minimum bending radius R (S104: affirmative). To do.

  Even when performing elliptic fitting using the above-described noise removal processing or the like, the cables can be accurately traced when the cables cross each other as in the above-described embodiment.

  In the above embodiment, the case where the feature point is the highest point of the ellipse has been described. However, the present invention is not limited to this, and the feature point may be a point that satisfies other predetermined conditions (for example, the center point of the ellipse). Good.

  The above processing functions can be realized by a computer. In that case, a program describing the processing contents of the functions that the processing apparatus should have is provided. By executing the program on a computer, the above processing functions are realized on the computer. The program describing the processing contents can be recorded on a computer-readable recording medium (except for a carrier wave).

  When the program is distributed, for example, it is sold in the form of a portable recording medium such as a DVD (Digital Versatile Disc) or a CD-ROM (Compact Disc Read Only Memory) on which the program is recorded. It is also possible to store the program in a storage device of a server computer and transfer the program from the server computer to another computer via a network.

  The computer that executes the program stores, for example, the program recorded on the portable recording medium or the program transferred from the server computer in its own storage device. Then, the computer reads the program from its own storage device and executes processing according to the program. The computer can also read the program directly from the portable recording medium and execute processing according to the program. Further, each time the program is transferred from the server computer, the computer can sequentially execute processing according to the received program.

  The above-described embodiment is an example of a preferred embodiment of the present invention. However, the present invention is not limited to this, and various modifications can be made without departing from the scope of the present invention.

In addition, the following additional remarks are disclosed regarding description of the above embodiment.
(Additional remark 1) The three-dimensional position of the point on the surface of the said cable corresponding to the lattice point in a two-dimensional surface from the image data which image | photographed the said cable in the state in which the cable was connected to some components of an electronic device An acquisition unit for acquiring data;
A setting unit that generates spatial data having a predetermined width in the extending direction of the cable in the vicinity of one end of the cable;
The cross-sectional shape data of the cable is generated from the point at which the three-dimensional position data existing in the spatial data is acquired, and the extending direction of the cable is specified based on the cross-sectional shape data and the cross-sectional shape data A processing unit that identifies a feature point that satisfies a predetermined condition, moves the spatial data based on the identified feature point and the extension direction, repeatedly executes a process, and generates the shape data of the cable from the processing result; ,
A control unit for causing the gripping device to grip the cable based on the generation result of the processing unit, and moving the gripping device to move the cable;
With
The processing unit uses the feature point and the extending direction to generate the shape data when the bending radius of the cable calculated based on the feature point of the specified cross-sectional shape data is less than a predetermined threshold. Without moving the space data a predetermined distance along the latest stretching direction that has not been discarded,
An electronic apparatus manufacturing apparatus characterized by the above.
(Appendix 2) The cable has a circular vertical cross-section,
The apparatus for manufacturing an electronic device according to appendix 1, wherein the predetermined distance is the same distance as a diameter of the vertical section.
(Additional remark 3) When the said bending radius is less than a predetermined threshold value, the said process part specifies the edge part of the cable which cross | intersects the said cable from the said image data, and only the distance based on the specified said edge part is said spatial data. The electronic apparatus manufacturing apparatus according to appendix 1, wherein the electronic apparatus is moved.
(Supplementary Note 4)
The feature point specified when the bending radius is equal to or greater than a predetermined threshold is stored in the storage unit,
Of the feature points stored in the storage unit, the bending radius of the cable is calculated based on two or more feature points stored immediately before and the newly specified feature points. The electronic device manufacturing apparatus according to any one of Supplementary notes 1 to 3.
(Additional remark 5) The three-dimensional position of the point on the surface of the said cable corresponding to the lattice point in a two-dimensional surface from the image data which image | photographed the said cable in the state in which the cable was connected to some components of an electronic device An acquisition unit for acquiring data;
A setting unit that generates spatial data having a predetermined width in the extending direction of the cable in the vicinity of one end of the cable;
The cross-sectional shape data of the cable is generated from the point at which the three-dimensional position data existing in the spatial data is acquired, and the extending direction of the cable is specified based on the cross-sectional shape data and the cross-sectional shape data A processing unit that identifies a feature point that satisfies a predetermined condition, moves the spatial data based on the identified feature point and the extension direction, repeatedly executes a process, and generates the shape data of the cable from the processing result; ,
A control unit for causing the gripping device to grip the cable based on the generation result of the processing unit, and moving the gripping device to move the cable;
With
When the bending radius of the cable calculated based on the specified feature point of the cross-sectional shape data is less than a predetermined threshold, the processing unit generates the feature point and the extending direction in generating the shape data. do not use,
An electronic apparatus manufacturing apparatus characterized by the above.
(Supplementary Note 6) The processing unit includes:
The feature point specified when the bending radius is equal to or greater than a predetermined threshold is stored in the storage unit,
Of the feature points stored in the storage unit, the bending radius of the cable is calculated based on two or more feature points stored immediately before and the newly specified feature points. The electronic device manufacturing apparatus according to appendix 5.
(Supplementary note 7) Acquire three-dimensional position data of points on the surface of the cable corresponding to lattice points in a two-dimensional plane from image data obtained by photographing the cable;
Generate spatial data having a predetermined width in the extending direction of the cable in the vicinity of one end of the cable,
The cross-sectional shape data of the cable is generated from the point at which the three-dimensional position data existing in the spatial data is acquired, and the extending direction of the cable is specified based on the cross-sectional shape data and the cross-sectional shape data Identify feature points that satisfy a predetermined condition, move the spatial data based on the identified feature points and the extension direction, repeatedly execute the process, and generate the shape data of the cable from the process result.
Let the computer execute the process,
In the process of generating the shape data of the cable, when the bending radius of the cable calculated based on the feature point of the specified cross-sectional shape data is less than a predetermined threshold, the feature point and the extending direction are It is not used for generation of shape data, and the spatial data is moved a predetermined distance along the latest stretching direction that has not been discarded.
A cable shape estimation program characterized by that.
(Appendix 8) The cable has a circular vertical cross-section,
The cable shape estimation program according to appendix 7, wherein the predetermined distance is the same distance as the diameter of the vertical section.
(Additional remark 9) In the process which produces | generates the shape data of the said cable, when the said bending radius is less than a predetermined threshold value, the edge part of the cable which cross | intersects the said cable is specified from the said image data, and the specified said edge part is The cable shape estimation program according to appendix 7, wherein the spatial data is moved by a distance based on the distance.
(Supplementary Note 10) In the process of generating the shape data of the cable,
The feature point specified when the bending radius is equal to or greater than a predetermined threshold is stored in the storage unit,
Of the feature points stored in the storage unit, the bending radius of the cable is calculated based on two or more feature points stored immediately before and the newly specified feature points. The cable shape estimation program according to any one of appendices 7 to 9.
(Additional remark 11) From the image data which image | photographed the cable, the 3-dimensional position data of the point on the surface of the said cable corresponding to the lattice point in a 2-dimensional surface are acquired,
Generate spatial data having a predetermined width in the extending direction of the cable in the vicinity of one end of the cable,
The cross-sectional shape data of the cable is generated from the point at which the three-dimensional position data existing in the spatial data is acquired, and the extending direction of the cable is specified based on the cross-sectional shape data and the cross-sectional shape data Identify feature points that satisfy a predetermined condition, move the spatial data based on the identified feature points and the extension direction, repeatedly execute the process, and generate the shape data of the cable from the process result.
Let the computer execute the process,
In the process of generating the shape data of the cable, when the bending radius of the cable calculated based on the feature point of the specified cross-sectional shape data is less than a predetermined threshold, the feature point and the extending direction are determined. Not used to generate the shape data,
A cable shape estimation program characterized by that.

10 Robot (gripping device)
16 Robot controller (control unit)
20 Cable 3D data acquisition unit (acquisition unit)
24 Search volume setting part (setting part)
26 Cable cross section / stretch direction calculation part (part of processing part)
28 Point of Interest Updater (part of the processor)
30 Cable shape estimation unit (part of processing unit)
36 Data storage (storage)
44 Cable 100 Electronic equipment manufacturing equipment

Claims (7)

3D position data of a point on the surface of the cable corresponding to a lattice point in a 2D plane is acquired from image data obtained by photographing the cable with a cable connected to a part of an electronic device component. An acquisition unit;
A setting unit that generates spatial data having a predetermined width in the extending direction of the cable in the vicinity of one end of the cable;
The cross-sectional shape data of the cable is generated from the point at which the three-dimensional position data existing in the spatial data is acquired, and the extending direction of the cable is specified based on the cross-sectional shape data and the cross-sectional shape data A processing unit that identifies a feature point that satisfies a predetermined condition, moves the spatial data based on the identified feature point and the extension direction, repeatedly executes a process, and generates the shape data of the cable from the processing result; ,
A control unit for causing the gripping device to grip the cable based on the generation result of the processing unit, and moving the gripping device to move the cable;
With
The processing unit uses the feature point and the extending direction to generate the shape data when the bending radius of the cable calculated based on the feature point of the specified cross-sectional shape data is less than a predetermined threshold. Without moving the space data a predetermined distance along the latest stretching direction that has not been discarded,
An electronic apparatus manufacturing apparatus characterized by the above. The cable has a circular vertical cross section,
The apparatus for manufacturing an electronic device according to claim 1, wherein the predetermined distance is the same distance as a diameter of the vertical section.   The processing unit specifies an edge portion of a cable that intersects the cable from the image data when the bending radius is less than a predetermined threshold, and moves the spatial data by a distance based on the specified edge portion. The apparatus for manufacturing an electronic device according to claim 1. The processor is
The feature point specified when the bending radius is equal to or greater than a predetermined threshold is stored in the storage unit,
Of the feature points stored in the storage unit, the bending radius of the cable is calculated based on two or more feature points stored immediately before and the newly specified feature points. The apparatus for manufacturing an electronic device according to any one of claims 1 to 3. 3D position data of a point on the surface of the cable corresponding to a lattice point in a 2D plane is acquired from image data obtained by photographing the cable with a cable connected to a part of an electronic device component. An acquisition unit;
A setting unit that generates spatial data having a predetermined width in the extending direction of the cable in the vicinity of one end of the cable;
The cross-sectional shape data of the cable is generated from the point at which the three-dimensional position data existing in the spatial data is acquired, and the extending direction of the cable is specified based on the cross-sectional shape data and the cross-sectional shape data A processing unit that identifies a feature point that satisfies a predetermined condition, moves the spatial data based on the identified feature point and the extension direction, repeatedly executes a process, and generates the shape data of the cable from the processing result; ,
A control unit for causing the gripping device to grip the cable based on the generation result of the processing unit, and moving the gripping device to move the cable;
With
When the bending radius of the cable calculated based on the specified feature point of the cross-sectional shape data is less than a predetermined threshold, the processing unit generates the feature point and the extending direction in generating the shape data. do not use,
An electronic apparatus manufacturing apparatus characterized by the above. Obtaining 3D position data of points on the surface of the cable corresponding to lattice points in a 2D plane from image data obtained by photographing the cable;
Generate spatial data having a predetermined width in the extending direction of the cable in the vicinity of one end of the cable,
The cross-sectional shape data of the cable is generated from the point at which the three-dimensional position data existing in the spatial data is acquired, and the extending direction of the cable is specified based on the cross-sectional shape data and the cross-sectional shape data Identify feature points that satisfy a predetermined condition, move the spatial data based on the identified feature points and the extension direction, repeatedly execute the process, and generate the shape data of the cable from the process result.
Let the computer execute the process,
In the process of generating the shape data of the cable, when the bending radius of the cable calculated based on the feature point of the specified cross-sectional shape data is less than a predetermined threshold, the feature point and the extending direction are It is not used for generation of shape data, and the spatial data is moved a predetermined distance along the latest stretching direction that has not been discarded.
A cable shape estimation program characterized by that. Obtaining 3D position data of points on the surface of the cable corresponding to lattice points in a 2D plane from image data obtained by photographing the cable;
Generate spatial data having a predetermined width in the extending direction of the cable in the vicinity of one end of the cable,
The cross-sectional shape data of the cable is generated from the point at which the three-dimensional position data existing in the spatial data is acquired, and the extending direction of the cable is specified based on the cross-sectional shape data and the cross-sectional shape data Identify feature points that satisfy a predetermined condition, move the spatial data based on the identified feature points and the extension direction, repeatedly execute the process, and generate the shape data of the cable from the process result.
Let the computer execute the process,
In the process of generating the shape data of the cable, when the bending radius of the cable calculated based on the feature point of the specified cross-sectional shape data is less than a predetermined threshold, the feature point and the extending direction are determined. Not used to generate the shape data,
A cable shape estimation program characterized by that.

Images