JP6170345B2 - Automatic correction system and method - Google Patents

Description

  The present invention relates to an automatic correction system and method, and more particularly to an automatic correction system and method for a welding robot using CAM data.

  In the welding process, when a plurality of the same products are produced, the installation position of the workpiece W is shifted every time. Therefore, a method of automatically correcting this shift using a CCD camera has been devised.

  See US Pat.

  On the other hand, when image processing is performed, the image of the target workpiece W often looks different due to surface scratches, shadows due to illumination, reflections, and the like. Therefore, each edge image is acquired and pattern matching is performed between the edge images. When a CAM is connected, pattern matching may be performed with the edge image of the workpiece W using the contour line output from the CAM as master information.

JP2011-45898

  When extracting an edge from an image, a false edge may be output due to various factors as described above. When pattern matching is performed using such an edge, erroneous detection is caused. Therefore, conventionally, the edge image is manually masked (removed), and the master information (the edge image of the welding point of the workpiece and the position information of the welding point is included, and welding of the second and subsequent workpieces W is performed. Information that is the basis of point detection). For this reason, man-hours for manually removing noise in the edge image and man-hours for instructing the position of the welding point are required, which increases the setup time and processing time of the operator, and during this time, the machine is completely out of the machine. I was unable to leave.

  Even when a CAM image (CAM data) is used, the contour line output from the CAM is used as an edge image of the master information as it is. The contour line output from the CAM image is an ideal edge. Actually, only two lines can be seen at the time of matching, and further, gaps, depth of field (lens depth of the lens) Since the degree of blurring of the line due to perspective) is not taken into consideration, there is a problem that the way the pattern matching target work W looks is considerably different.

  SUMMARY OF THE INVENTION An object of the present invention is to create master information for pattern matching without using a CAM image, and to detect welding points by pattern matching and automatically correct a welding robot. .

The present invention is for solving the above-mentioned problems, and the invention according to claim 1 is an automatic correction system for a welding robot , wherein a camera capable of photographing a welding point of a workpiece, the camera is controlled, and a workpiece image is obtained. An edge is extracted from an imaging unit that acquires the image, a CAM that creates the appearance of the camera in a virtual space and outputs it as a CAM image, and a welding point image of a workpiece photographed by the camera and the imaging unit. An edge extraction unit that creates an image; a welding point position information acquired by the CAM; and a master information creation unit that creates master information at the welding point of the workpiece from the edge image acquired by the edge extraction unit; wherein using the master information created by the master information creation unit, a detecting unit for detecting the position of the welding points of the same workpiece to be machined in the second and later, the pattern It means for automatically corrected by the correction amount that has been generated based on the detection of the position by etching, characterized by comprising a.

The invention according to claim 2 is characterized in that the edge extraction unit extracts an edge from a welding point image of a workpiece by a linear filter .

  The invention according to claim 3 is characterized in that an expanded CAM image is created by expanding a predetermined line included in the CAM image by a predetermined amount.

According to a fourth aspect of the present invention, in the automatic correction method for a welding robot, a process of acquiring a workpiece image by controlling a camera capable of photographing a welding point of a workpiece, and creating an appearance on the virtual space in the virtual space And outputting as a CAM image; extracting a edge from a welding point image of a workpiece photographed by the camera; creating an edge image; welding point position information based on the CAM image; Generating based on the step of creating master information at the welding point, detecting the position of the welding point of the same workpiece to be processed after the second using the master information, and detecting the position by pattern matching And a step of performing automatic correction with the corrected amount.
The invention according to claim 5 is characterized in that the edge extraction unit extracts an edge from a welding point image of a workpiece by a linear filter.

  The present invention is a system including a CCD camera, an imaging unit, a CAM, an edge extraction unit, a master information creation unit, and a detection unit. Then, the workpiece W is photographed by the CCD camera and the imaging unit, the master information is prepared by the CAM, the edge extraction unit, and the master information creation unit, and the welding point position of the workpiece W is detected by the detection unit.

  Thereby, the master information is automatically created by the first workpiece W, and the positions of the welding points of the second and subsequent workpieces W can be automatically corrected.

  And it becomes possible to correct the set deviation of the workpiece W on the jig automatically and accurately without bothering the operator's hand, and the operator can save the time and effort required for the operation. The effect is that it can be directed to other work.

It is the schematic which shows the outline of the automatic correction system of a welding robot. It is a flowchart which shows operation | movement of the automatic correction system of a welding robot. It is a flowchart which shows operation | movement of the automatic correction system of a welding robot. It is explanatory drawing explaining preparation of a master image.

  Embodiments of the present invention will be described with reference to the drawings.

  The main configuration of the automatic correction system S will be described with reference to FIG.

  The automatic correction system S includes means for using the captured image and the CAM image (CAM data), creating master information for pattern matching, and performing automatic correction based on the pattern matching.

  Pattern matching in this example refers to searching for matching positions of welding points by comparing images.

    This will be described in detail. The automatic correction system S is mounted on the tip of the welding head H and the like, and can capture a welding point of the workpiece W. The CCD camera 1 controls the CCD camera 1 and acquires an image of a necessary welding point. Part 2 and the appearance of the CCD camera 1 are created in a virtual space, and an image (which is output as a wire frame subjected to hidden line processing and represents an ideal edge for the appearance of the CCD camera 1) Since it also has position information of the welding point to be output, from the CAM3 that outputs where the welding point is on the image and the welding point image of the workpiece W taken by the CCD camera 1 and the imaging unit 2, an edge is obtained by a linear filter. Are extracted from the edge extraction unit 4 that creates the edge image b, the welding point position information acquired by the CAM 3, and the edge image b acquired by the edge extraction unit 4. The master information creation unit 5 that creates the master information g at the welding point of the workpiece and the master information g created by the master information creation unit 5, the welding points of the same workpiece W to be processed after the second The detection unit 6 detects the position.

  In this example, the welding point position is the coordinates of the position at which welding is performed, and includes the coordinates of the welding start point and end point.

  The robot R includes a welding head H, and the arm structure of the robot arm is, for example, six axes. A swivel axis that rotates the body (S axis), a lower arm axis that moves the body back and forth (L axis), an upper arm axis that moves the arm up and down (U axis), a wrist swivel axis that rotates the arm (R axis), It comprises a wrist bending axis (B axis) that swings up and down, and a wrist rotation axis (T axis) that rotates the wrist.

  The automatic correction system S includes a computer, and includes a main body (including a central processing unit (CPU), a random access memory (RAM), a read only memory (ROM)), an input / output unit (a display unit, a keyboard, and the like). ).

  In the present embodiment, control of an industrial robot will be described as an example, but the present invention is not particularly limited to an industrial robot.

  The operation of the automatic correction system S configured as described above will be described.

  Please refer to FIG. In step SA01, an image of the first workpiece W is taken to create a workpiece image a.

  In step SA02, the workpiece image a is filtered to extract an edge to create an edge image b.

  In step SA03, the CAM 3 outputs a CAM image c (a wire frame image subjected to hidden line processing that can be assumed to be viewed from the same position as the CCD camera) and a welding point position d.

  In step SA04, the line of the CAM image c is expanded by a predetermined amount to create an expanded CAM image e.

  In steps SA05 and SA06, the noise of the edge image b is removed, and the final edge image f is created. That is, the edge image b and the expanded CAM image e are merged, a portion outside the expanded line of the expanded CAM image e is masked (removed), and the edge image b is output as the final edge image f.

  In step SA07, the final edge image f and the welding point position d are combined and stored as master information g.

  As described above, the master information g is automatically created from the first work W.

  Next, a procedure for automatically correcting the second and subsequent workpieces W by the automatic correction system S will be described.

  In step SB01, an image of the second (and subsequent) work W is taken to create a work image a.

  In step SB02, the workpiece image a is filtered to extract an edge to create an edge image b.

  In step SB03, master information g stored at the time of the first work is acquired.

  In step SB04, pattern matching is performed on the edge image b and the edge image of the master information g, and the position on the edge image b where the welding point position of the master information g is detected is detected.

  In step SB05, the amount of deviation between the center of the current image and the detected point is the correction amount. Then, correction is performed based on this correction amount.

  Although it has been described that a filter is used when performing edge extraction as described above, any method of edge extraction may be used.

  The merge of the work image a and the CAM image c and creation of the master information g will be described in detail with reference to FIG.

  On the screen G1, as the edge image b extracted from the work image a, the plate IT1 and the plate IT2 are displayed in the gap S. A scratch K is displayed on the plate IT1, and a light pattern M is displayed on the plate IT2. A fracture line HL2 is displayed near the center of the plate thickness displayed from HL1 to HL3.

  On the screen G2, a line L1, a line L2, and a line L3 are displayed as the CAM image c. On the screen G3, an expanded CAM image e obtained by expanding the CAM image c is displayed. In other words, an expanded CAM image e including the lines BL1, BL2, and BL3 obtained by expanding the lines L1, L2, and L3 to the width B dimension is displayed.

  An image obtained by merging the edge image a and the expanded CAM image e is displayed on the screen G4.

  The edge image a included in the range of the expanded CAM image e (the range indicated by the B dimension) is left. As a result, as shown in the screen G5, the final edge image f including the lines HL1, HL3 and the plates IT1 and IT2 partitioned by the gap S is obtained.

  The present invention is not limited to the embodiments of the invention described above, and can be implemented in other modes by making appropriate modifications.

S Automatic correction system 1 CCD camera 2 Imaging unit 3 CAM
4 Edge extraction unit 5 Master information creation unit 6 Detection unit R Robot H Welding head

Claims (5)

In the automatic correction system for welding robots,
A camera capable of photographing a welding point of a workpiece, an imaging unit for controlling the camera and acquiring a workpiece image;
A CAM that creates the appearance of the camera in a virtual space and outputs it as a CAM image;
An edge extraction unit that extracts an edge and creates an edge image from a welding point image of a workpiece photographed by the camera and the imaging unit;
A master information creation unit that creates master information at the welding point of the workpiece from the welding point position information obtained by the CAM and the edge image obtained by the edge extraction unit;
Using the master information created by the master information creation unit, a detection unit for detecting the position of the welding point of the same workpiece to be processed after the second,
An automatic correction system comprising: means for performing automatic correction with a correction amount generated based on position detection by pattern matching. The automatic correction system according to claim 1, wherein the edge extraction unit extracts an edge from a weld point image of a workpiece by a linear filter.   The automatic correction system according to claim 1, wherein an expanded CAM image is generated by expanding a predetermined line included in the CAM image by a predetermined amount. In the automatic correction method for welding robots,
A process of acquiring a workpiece image by controlling a camera capable of photographing a welding point of the workpiece;
Creating the appearance of the camera in a virtual space and outputting it as a CAM image;
From the welding point image of the work photographed by the camera, to extract an edge and create an edge image;
Creating welding point position information by the CAM image and master information at the welding point of the workpiece from the edge image;
Using the master information to detect the position of the welding point of the same workpiece to be processed after the second,
And automatic correction with a correction amount generated based on position detection by pattern matching. The automatic correction method according to claim 4, wherein the edge extraction unit extracts an edge from a weld point image of a workpiece by a linear filter.

Images