JP5147255B2 - Image processing verification system - Google Patents

Description

  The present invention relates to an image processing verification system that reproduces and verifies image processing using image processing reproduction data stored in a database.

In various production apparatuses such as electronic component mounting machines in recent years, as described in Patent Document 1, the role of image processing is large, and various purposes (for example, recognition / positioning of parts / marks, inspection, height) Measurement). Because of its wide range of applications, it is necessary for the operator to appropriately specify various settings for image processing according to the purpose. In order to verify the quality of these settings, a trial-and-error (trial and error) method is used. In addition, it was necessary to repeat the verification process of confirming the quality of various settings of image processing by repeating the production process many times. Moreover, it is necessary to redo the verification work every time various settings of the image processing are changed.
Japanese Patent No. 2698258

  The conventional verification method has a drawback that it takes a lot of time for the verification work because it is necessary to actually repeat the production process many times by the trial and error method. In addition, for example, even if the imaging object is changed or the imaging object is the same, the image processing conditions are not the same because the imaging conditions of the camera are changed. In addition, there is a problem that it is difficult to perform reproduction verification under the same conditions by changing various image processing conditions.

  Recently, a solution to support verification work has been developed by providing an offline image processing device (image processing application that runs on a personal computer) separately from the actual machine, but this solution is also basically Since it is necessary to repeat the same trial and error as the actual machine, it was difficult in terms of time to fully verify the huge combination of conditions and settings.

  Also, with conventional electronic component mounters, when image processing is performed, image processing can be reproduced only with the image data that was last subjected to image processing or the image data remaining in the buffer of the electronic component mounter. Therefore, the reproducible image data is very small and sufficient verification work cannot be performed. Moreover, normally, since the process proceeds to the next cycle unless the production of the actual machine is stopped due to an image processing error or the like, it is not possible to reproduce a defect (such as an angular deviation) when no image processing error occurs.

  Also, if the image processing was successful, but data different from the actual angle (coordinates) was obtained as the image processing result, the actual parts were mounted on a trial basis, and the results were aggregated. Thus, it is necessary to carry out a verification work of confirming the validity of the image processing conditions (images, part data, etc.), and the verification work takes an enormous amount of time. Moreover, if actual production is started without verification, a defective product will be produced. Furthermore, when image processing conditions (images, part data, etc.) are changed, it is necessary to testly mount using actual parts in order to verify whether the conditions are correct.

The present invention has been made in consideration of these circumstances. Therefore, the object of the present invention is to accurately reproduce the image processing executed by the image processing apparatus of the electronic component mounting machine and accurately verify the effectiveness of the image processing conditions. It is an object of the present invention to provide an image processing verification system that can easily implement the above.

In order to achieve the above object, the invention according to claim 1 stores data that can reproduce image processing executed by the image processing apparatus of the electronic component mounting machine (hereinafter referred to as “image processing reproduction data”) as a database. Image processing reproduction data creating / saving means and a database stored in the image processing reproduction data creating / saving means obtain image processing reproduction data necessary for image processing to be verified and use the image processing reproduction data to obtain an image. An image processing verification unit that reproduces the processing and verifies the effectiveness of the image processing conditions, and the image processing reproduction data creation / storing unit performs image processing in the image processing apparatus and performs actual production. In addition, when the image processing conditions match the specified conditions, at least the captured image data is saved as image processing reproduction data. That.

In this configuration, it is possible to verify the effectiveness of the image processing conditions by reproducing the image processing using the image processing reproduction data stored in the database. It is no longer necessary to repeat the above process many times, and accurate verification work can be easily performed. In addition, since the verification work can be performed using an offline image processing apparatus, accurate verification work can be easily performed offline without using an actual machine. Furthermore, when upgrading the firmware of the actual machine, there is an advantage that the user side can easily carry out original verification in addition to the manufacturer verification.

In the present invention, among the data used in the image processing apparatus of the electronic component mounting machine, at least the captured image data is made into a database as image processing reproduction data. For example, the part data of the component is changed. In this case, it is possible to easily verify whether or not image processing can be performed on the past images stored in the database without any problem. In addition, when image processing is performed by the image processing apparatus and actual production is performed, if the image processing conditions match the specified conditions, at least captured image data is stored as image processing reproduction data. Thus, there is an advantage that the image processing reproduction data can be automatically stored in the database during actual production.

In this case, as described in claim 2 , the image processing verification means uses part data, mark data, camera images, imaging conditions, and camera resolution used by the user in reproducing the image processing. Change at least one of camera distortion correction value, sequence number, device number, execution time, firmware version, mounting machine name, X coordinate value, Y coordinate value, Q angle, and error code of the image processing result A configuration provided with means is preferable. Thereby, original verification can be easily implemented on the user side.

  However, according to the present invention, an arbitrary database may be created on the user side, or the contents of an existing database may be duplicated / changed to create a database in which various conditions and settings are changed. Also good.

Hereinafter, an embodiment in which the present invention is applied to an electronic component mounting machine will be described.
The image processing verification system of the present embodiment is configured using an offline image processing apparatus 11 of an electronic component mounting machine. The offline image processing apparatus 11 images an input device 12 such as a keyboard and a mouse, a display 13 that displays image processing results and verification results, and parts (parts) sucked by a suction nozzle of an electronic component mounting machine. And a storage device 15 such as a hard disk drive is provided internally or externally.

  In this embodiment, the off-line image processing apparatus 11 functions as image processing reproduction data creation / storage means for storing data that can reproduce image processing (hereinafter referred to as “image processing reproduction data”) as a database and storing the data in the storage device 15. . Note that the image processing reproduction data database may be stored in a network server.

  Further, the off-line image processing apparatus 11 retrieves necessary image processing reproduction data from the database stored in the storage device 15 and reproduces and verifies the image processing using the image processing reproduction data. In addition to functioning, when reproducing image processing, the user has a function of changing conditions and settings of the image processing by operating the input device 12.

  In this embodiment, the off-line image processing apparatus 11 performs image processing reproduction data on image data that satisfies the specified image processing conditions and parameters necessary for the image processing during actual production by the electronic component mounting machine. Is stored in the storage device 15 (or network server). It is also possible to change the image processing condition parameters in the stored data and load and unload them to the electronic component mounter. The image processing reproduction data can be further narrowed down by filtering according to the search conditions. It is like that.

Here, as the storage conditions of the image processing reproduction data (combinations of conditions are possible), for example, the following conditions can be cited.
(1) The difference between the image processing result and the expected value is greater than the specified value
(2) With specified part name
(3) For the specified device number (feeder placement position)
(4) For the specified sequence number
(5) An image processing error has occurred
(6) With specified image processing algorithm
(7) Specified mark type
(8) With specified mark name
(9) Specified nozzle type
(10) For the specified nozzle number

Moreover, as a search condition, the following conditions are mentioned, for example.
(1) Specified nozzle type
(2) For the specified nozzle number
(3) with specified image processing algorithm
(4) With specified part name
(5) With specified mark name
(6) For the specified camera
(7) For the specified production program
(8) For the specified device number
(9) For the specified sequence number
(10) With specified image processing error code

Examples of parameters necessary for image processing include the following.
(1) Shape data (part data, mark data)
(2) Camera image
(3) Imaging conditions (nozzle type and light source type)
(4) Camera resolution
(5) Calibration data (camera distortion correction value)
(6) Image processing conditions (sequence number, device number, execution time, firmware version, collected machine name)
(7) Image processing results (X coordinate value, Y coordinate value, Q angle, error code, etc.)

  In this embodiment, various image processing conditions are set at the time of production, and image processing reproduction data matching the various conditions is stored. As a result, it can be used for the reduction of defective products and the evaluation of image processing conditions (part data, etc.), the appropriate image processing conditions can be tested, and productivity can be increased.

  For example, if there is a part that rarely causes an angle shift, by setting the sequence number or allowable angle of that part as a condition, the image that matches the condition and the image processing condition parameter are saved, and the image processing is performed. Reproducibility makes it easy to solve problems.

  Further, for example, by changing various conditions, loading the electronic component mounting machine and performing re-image processing, it becomes easy to verify the effectiveness of the changed image processing conditions, and productivity can be improved.

  Next, a procedure for reproducing and verifying image processing using image processing reproduction data will be described with reference to the flowcharts of FIGS.

[Normal test case]
The flowchart of the normal test case shown in FIG. 2 shows a procedure for testing the ability to recognize normal image processing data (shape data) as “normal”. When processing this normal test case, it is preferable to prepare a plurality of lots and vendor parts.

  In the normal test case, first, shape data is created (step 101). This shape data creation method is, for example, extracting a part shape from a drawing etc. with an editor and inputting the numerical value as shape data, or picking up an image of the part sucked by the suction nozzle of the electronic component mounting machine with a camera. There is a method of setting conditions for image processing by an operator and using the data obtained by the image processing as shape data.

  Thereafter, a database that expects n normal image processing results is prepared (step 102). This database may be prepared by setting various image processing conditions during production and storing image processing reproduction data that matches the various conditions.

  Then, n times of image processing tests are executed using a database that expects n normal image processing results (step 103). This image processing test is executed by an image processing application installed in the offline image processing apparatus 11. The image processing test may be executed by a computer that controls the electronic component mounting machine.

  If an image processing error occurs contrary to expectations in this image processing test, it is determined that the shape data used for the image processing is inappropriate, the shape data is corrected (step 104), and the corrected shape data is obtained. The image processing test is executed again using the data.

  If the shape data is changed or the firmware is upgraded, the image processing result (X, Y, Q) will change. If the amount of change in the image processing result exceeds the allowable value, the mounting accuracy will be adversely affected. Is determined to be an image processing error.

  As described above, the database that expects n normal image processing results is inadequate shape data until normal image processing results are obtained as expected (that is, until no image processing error occurs). The process of re-executing the image processing test with corrections is repeated, and when the normal image processing result is obtained as expected for all databases that expect n normal image processing results, the normal system test is passed Is determined.

[Abnormal test case]
The flowchart of the abnormal test case shown in FIG. 3 shows a procedure for testing the ability to recognize abnormal image processing data (shape data) as “abnormal”. In this abnormal test case, for example, it is confirmed that various tolerance checks work effectively, or that an image processing error occurs when components are crossed.

In the abnormal test case, first, shape data is created (step 201).
Thereafter, a database that expects n image processing errors is prepared (step 202). Then, n image processing tests are executed using a database that expects n image processing errors (step 203).

  In this image processing test, if the image processing result is “normal” contrary to expectations (that is, if no image processing error occurs as expected), the shape data used for the image processing is inappropriate. The shape data is corrected (step 204), and the image processing test is executed again using the corrected shape data.

  As described above, with respect to a database that expects n image processing errors, the process of correcting inappropriate shape data and re-executing the image processing test is repeated until an image processing error is obtained as expected. As for all databases that expect n image processing errors, it is determined that the abnormal system test has passed when all the image processing errors are obtained as expected.

[Multi-vendor support]
The flowchart corresponding to a plurality of vendors shown in FIG. 4 shows a procedure for correcting the shape data as necessary when using parts of different vendors. In this multi-vendor procedure, first, a test case for a part of Company A is prepared from a database or the like (step 301). Thereafter, the shape data is corrected for the parts of company B (step 302).

  Then, a test case for parts of Company B is prepared from a database or the like (step 303). Thereafter, an image processing test is performed on all test cases (step 304). As a result, if there is shape data that has failed the image processing test, the shape data is corrected (step 305). ) Perform the image processing test again.

  As described above, in all test cases, the process of correcting inappropriate shape data and re-executing the image processing test is repeated until the result of the image processing test is passed. When the result of the image processing test is passed, this processing is terminated.

[Upgrade verification]
The flowchart for version upgrade verification shown in FIG. 5 shows a procedure for performing original verification on the user side when the firmware is upgraded. In the upgrade verification procedure, first, a test case is prepared from a database or the like (step 401). Thereafter, the firmware is upgraded (step 402).

  Thereafter, an image processing test is performed on all test cases (step 403), and as a result, if there is shape data in which the result of the image processing test fails, whether or not there is a defect in the new version. Is determined based on an upgrade report or the like (step 404). At this time, for example, it is determined whether or not there is a defect in the new version based on whether or not the cause of the failure is due to a change intended by the version upgrade.

  As a result, if it is determined that the new version has a problem, the firmware is returned to the original version (step 406). On the other hand, if it is determined that there is no defect in the new version, the shape data whose image processing test result is rejected is corrected (step 405), and the image processing test is performed again.

  As described above, if there is no defect in the new version, correct the inappropriate shape data and re-execute the image processing test until the image processing test result passes in all test cases. This process is terminated when the result of the image processing test passes in all test cases.

It is a block diagram which shows schematic structure of the image processing verification system in one Example of this invention. It is a flowchart of a normal system test case. It is a flowchart of an abnormal system test case. It is a flowchart corresponding to multiple vendors. It is a flowchart of version-up verification.

Explanation of symbols

  11: Offline image processing device (image processing reproduction data creation / storage means, image processing verification means), 14 ... Camera, 15 ... Storage device

Claims (2)

Image processing reproduction data creation / storage means for storing data that can reproduce image processing executed by the image processing apparatus of the electronic component mounting machine (hereinafter referred to as “image processing reproduction data”) as a database;
Image processing reproduction data necessary for the image processing to be verified is acquired from the database stored in the image processing reproduction data creation / storage unit, and the image processing is reproduced using the image processing reproduction data to validate the image processing conditions. Image processing verification means for verifying
The image processing reproduction data creation / storage means performs image processing by the image processing apparatus and performs actual production, and if the image processing conditions match specified conditions, at least the captured image An image processing verification system characterized by storing the data as image processing reproduction data .   The image processing verification means is a part data used by the user when reproducing the image processing, mark data, camera image, imaging conditions, camera resolution, camera distortion correction value, sequence number, It has means for changing at least one of a device number, execution time, firmware version, mounting machine name, X coordinate value, Y coordinate value, Q angle, and error code of the image processing result. The image processing verification system according to claim 1.

Images