An automatic teaching method for visual inspection of glue coating defects

Abstract

The present invention discloses an automatic teaching method for visual detection of glue defects, which relates to the technical field of visual detection of glue defects in intelligent manufacturing. The method aims to solve the problems of low automation level of existing equipment program teaching, manual debugging required during equipment use, long equipment deployment time and inflexible use, high technical requirements for operators, and high labor costs. Step 1: Install the collector on the glue robot and start the robot to complete a glue application; Step 2: Set basic parameters and set the project of visual detection of glue defects through the interactive interface; Step 3: The user clicks one-button teaching through the interactive interface, and the program automatically controls the collector to change the color of the fill light and collect image photos at a fixed time, filters, synthesizes, binarizes, compares standards, identifies features, calculates detection project parameters, etc., and starts the automatic recognition task; Step 4: After the automatic recognition task is completed, the final result is displayed through the status display on the interactive interface.

Description

Technical Field

[0001] The present invention relates to the technical field of visual detection of gluing defects in intelligent manufacturing, and specifically to an automatic teaching method for visual detection of gluing defects. Background Art

[0002] In the field of visual inspection of glue coating defects in intelligent manufacturing, the working principle of the detection method or device is to capture images and use image processing algorithms to obtain actual working parameters based on parameters such as glue, gluing standards and external environment. These parameters, such as glue color, glue width, light color, etc., are then compared with the set parameters to determine the completion quality of the gluing task.

[0003] Nowadays, a common method or device is for technical R&D personnel to solidify relevant parameters such as glue, collector, and external environment in the running program, and does not provide the end user with a human-machine interface for changing parameters. This shows that for specific implementation projects, technical personnel need to perform on-site debugging. When the glue detection work task changes, the existing program cannot meet the usage requirements and needs to be reprogrammed and debugged accordingly. This method or device takes a long time to deploy, is inflexible to use, has high technical requirements for operators, and the labor cost of each task change is high. Products that provide teaching interfaces have appeared on the market. They parameterize the working process of glue tasks, glue coating equipment, and collectors, and write human-machine operation interfaces and operating instructions. Users input working parameters into the running detection program through the human-machine operation interface. These parameters include but are not limited to: working distance, fill light color, glue color, base color, pixel size. There is a conversion relationship between the parameters, which needs to be solved by the user. In addition, the user is required to mark the position reference on the collected image. These operations facilitate product deployment and change tasks, but the professional and technical requirements of the integration operator are still high. Summary of the Invention

[0004] The purpose of the present invention is to provide an automatic teaching method for visual detection of glue coating defects, so as to solve the problem raised in the above background technology that the existing equipment program cannot meet the use needs, manual debugging is required during the use of the equipment, and the equipment deployment time is long and inflexible to use, and the technical requirements for operators are high, thus resulting in high labor costs.

[0005] To achieve the above object, the present invention provides the following technical solution: an automatic teaching method for visual inspection of glue coating defects, comprising a collector, a process module and an interactive interface,

[0006] Step 1: Install the collector on the gluing robot and start the robot to complete one gluing operation;

[0007] Step 2: Set basic parameters through the interactive interface and set the items for visual inspection of glue defects;

[0008] Step 3: Through the interactive interface, the user clicks one-button teaching, and the program automatically controls the collector to change the fill light color and collect images, and then performs filtering, synthesis, binarization, comparison standards, feature recognition, and calculation of detection project parameters on the image to start the automatic recognition task;

[0009] Step 4: After the automatic recognition task is completed, the final result is displayed through the status display on the interactive interface;

[0010] Step 5: The user determines whether the marked range is correct. If it is correct, click the [Confirm] button on the interactive interface. If it is incorrect, click the [Cancel] button to re-identify.

[0011] Step 6: Perform one-key teaching (9) for each selected qualified glue coating section in turn. After the teaching is completed, press the [Finish] key to perform weighted evaluation on all teaching results, and the sum of the weights is equal to 1. The glue coating defect visual inspection task is stored and the teaching is completed.

[0012] By adopting the above technical solution, based on the combination of collector, process module and interactive interface, software programs and algorithms are used to automatically identify the user's expectations for defect detection. General technicians only need to base their work on their skills and experience.

[0013] Furthermore, the collector consists of a glue gun, three collection cameras and a collection fill light. The process module includes setting basic parameters, one-click teaching, automatic identification of user expectations, user confirmation and task issuance. The collector is electrically connected to the process module. The interactive interface is provided with a basic parameter area, one-click teaching, identification display and status display. The interactive interface includes a completion button. The identification display shows the image status of the glue strip, marking box and glue gun.

[0014] By adopting the above technical solution, it can be suitable for users who have no knowledge and skills related to image processing. During use, users do not need to set parameters related to image processing, such as pixel size ratio, shooting distance, etc., and do not need to perform operations related to image processing, such as marking reference points on the image. They only need to set parameters related to the gluing task, such as task number, defect detection items, etc.

[0015] Furthermore, the identification display includes three acquisition cameras, a glue gun, a marking frame and a glue strip.

[0016] By adopting the above technical solution, the position images between the three acquisition cameras, the glue gun, the marking frame and the glue strip in the current state are identified and displayed.

[0017] Furthermore, the set basic parameters include parameters marking the characteristics of the work task, such as the number of teaching times, work task number, and component number, and the completion button can be used to calculate the weighted average according to the results of each teaching.

[0018] By adopting the above technical solution, one-key teaching is started. One teaching includes all the teaching processes. Each teaching requires pressing one-key teaching once. After completing the teaching, press the completion button. The weighted average is calculated according to the results of each teaching. The content of the detection item, such as size and distance, does not need to set specific size and distance. One-key teaching is a key command input. The user confirms that the teaching preparation is completed through this command and starts the automatic recognition task. The specific usage process is: 1. In the automatic teaching interface, the user sets the defect detection item, which is size, margin, or one of them. 2. Move the collector to the qualified glue coating section in the working posture. The computer program automatically controls the collection of fill light to illuminate the glue coating section and take samples. 3. The program automatically identifies the glue strip and calculates the pixel width. After successful recognition, the result is output and displayed, and the qualified glue strip section is marked. 4. The user observes the result to confirm whether it is consistent with the expected defect detection standard.

[0019] Furthermore, the automatic identification of user expectation is that the program automatically identifies the user's detection expectation based on the collected images.

[0020] By adopting the above technical solution, it is applicable to users who have no knowledge and skills related to image processing, and the users do not need to set parameters related to image processing during use.

[0021] Furthermore, the above-mentioned recognition content includes adjusting the color of different fill lights, collecting images, image preprocessing, cutting pictures, identifying features, quantizing grayscale values ​​from 0 to 255, and calculating pixel values ​​of straight line width.

[0022] By adopting the above technical solution, taking a 3-camera collector as an example, then:

[0023] 1) Adjust the color of different fill lights and take photos. If the subsequent photos cannot meet the recognition requirements, the sequence is set to R, G, B, RGB three-color step combination. If the combination is completed and still cannot complete the task, terminate;

[0024] 2) collecting images;

[0025] 3) Image preprocessing, including filtering, image synthesis and grayscale conversion;

[0026] 4) Crop the image and identify features, looking for straight line features in the affine direction outward from the center of the image and for larger, more uniform areas outside the straight lines. If no straight line features are identified, return to step 1).

[0027] 5) Quantize the grayscale value from 0 to 255 and compare the grayscale difference between the line feature and the surrounding area. If the grayscale difference is ≥ 128, proceed to the next step, otherwise return to step 1);

[0028] 6) Calculate the pixel value of the straight line width to automatically identify the user's expectations.

[0029] Furthermore, the one-key teaching user confirms through this command that the teaching preparation is complete and starts the automatic recognition task.

[0030] By adopting the above technical solution, the final result is more accurate.

[0031] Furthermore, the one-key teaching includes displaying the identified user expectations on the interface, marking the rubber strip contour, edge, etc., and the user confirms that the marked rubber strip contour, edge, etc. information meets the user expectations.

[0032] By adopting the above technical solution, the user can judge whether the mark on the interface image is a rubber strip. If it is satisfied, the user can confirm and proceed to the next step. If not, the user can repeat the teaching.

[0033] Furthermore, the task delivery includes saving and automatically identifying user expectations and delivering related work software modules.

[0034] By adopting the above technical solution, the equipment and system can be controlled.

[0035] Furthermore, the collection fill light is an RGB fill light.

[0036] By adopting the above technical solution, the fill light is collected as an RGB three-color light, and the light color is R, G, B three colors and any combination thereof.

[0037] Compared with the prior art, the present invention has the following beneficial effects:

[0038] 1. This automatic teaching method for visual inspection of gluing defects is suitable for users who do not have knowledge and skills related to image processing. During use, users do not need to set parameters related to image processing, such as pixel size ratio, shooting distance, etc., and do not need to perform operations related to image processing, such as marking reference points on the image. They only need to set parameters related to the gluing task, such as task number, defect detection items, etc.

[0039] 2. The automatic teaching method of visual inspection of gluing defects can reduce the need for professional knowledge and skills of the user population. It uses software programs and algorithms to automatically identify the user's expectations for defect detection. General technicians only need to set the work tasks of the visual inspection equipment for gluing defects based on their job skills and experience. No image-related professional knowledge is required. It is easy to use and suitable for a wide range of people. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 It is a schematic diagram of the structure of the collector of the present invention;

[0041] Figure 2 A schematic diagram of the user interaction interface of the present invention;

[0042] Figure 3 It is a schematic diagram of an implementation flow chart of the present invention;

[0043] Figure 4 This is a schematic diagram of the automatic identification display area of ​​the present invention;

[0044] Figure 5 This is a schematic diagram showing the identification of the present invention;

[0045] Figure 6 It is a schematic diagram of the process module of the present invention.

[0046] In the figure: 1. Collector; 2. Glue gun; 3. Collection camera; 4. Collection fill light; 5. Marking box; 6. Glue strip; 7. Process module; 8. Setting basic parameters; 9. One-click teaching; 10. Automatic recognition of user expectations; 11. User confirmation; 12. Task issuance; 13. Interactive interface; 14. Basic parameter area; 16. Identification display; 17. Status display. DETAILED DESCRIPTION

[0047] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. The described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0048] An automatic teaching method for visual detection of glue coating defects, the present invention provides the following technical solutions:

[0049] like Figure 1-6 As shown,

[0050] Step 1: Install the collector 1 on the gluing robot and start the robot to complete one gluing operation;

[0051] Step 2: Set the basic parameters 8 through the interactive interface 13 and set the items for visual inspection of glue defects;

[0052] Step 3: The user clicks on the one-button teaching 9 through the interactive interface 13, and the program automatically controls the collector 1 to change the fill light color and collect image photos, and performs filtering, synthesis, binarization, comparison standards, feature recognition, calculation of detection project parameters, etc. on the image, and starts the automatic recognition task;

[0053] Step 4: After the automatic recognition task is completed, the final result is displayed through the status display 17 on the interactive interface 13;

[0054] Step 5: The user determines whether the marked range is correct. If it is correct, click the [Confirm] button on the interactive interface 13; if it is incorrect, click the [Cancel] button to re-identify;

[0055] Step 6: Perform one-key teaching (9) for each selected qualified glue coating section in turn. After the teaching is completed, press the [Finish] key to perform weighted evaluation on all teaching results, and the sum of the weights is equal to 1. The glue coating defect visual inspection task is stored and the teaching is completed.

[0056] The collector 1 is composed of a glue gun 2, three collection cameras 3 and a collection fill light 4. The process module 7 includes setting basic parameters 8, one-key teaching 9, automatic recognition of user expectations 10, user confirmation 11 and task issuance 12. The collector 1 is electrically connected to the process module 7. The interactive interface 13 is provided with a basic parameter area 14, one-key teaching 9, recognition display 16 and status display 17. The interactive interface 13 includes a completion button. The recognition display 16 includes three collection cameras 3, a glue gun 2, a marking frame 5 and a glue strip 6. The basic parameters 8 include the number of teaching times, work task number, part number and other parameters for marking work task characteristics. The completion button can be used to calculate the weighted average according to the results of each teaching. Automatic recognition of user expectations 10 is the expectation of automatic recognition of user detection by the program based on the collected image. Based on the above recognition content, it includes adjusting the color of different fill lights, collecting images, Image preprocessing, image cutting, feature recognition, quantization of grayscale values ​​with 0 to 255 and calculation of straight line width pixel values, one-click teaching 9. The user confirms through this command that the teaching preparation is complete and starts the automatic recognition task. One-click teaching 9 includes displaying the recognized user expectations on the interface, marking the rubber strip contour, edge, etc., user confirmation 11 is to confirm that the marked rubber strip contour, edge and other information meet the user's expectations, task delivery 12 includes saving automatic recognition user expectations 10, issuing related work software modules, collecting fill light 4 as RGB fill light, which can reduce the need for professional knowledge and professional skills of the user population, and using software programs and algorithms to automatically identify user defect detection expectations. General technicians only need to set the work tasks of the glue defect visual inspection equipment according to their job skills and experience. They do not need image-related professional knowledge, are simple to use, and are suitable for a wide range of people.

[0057] Working principle: Combination Figure 1 、 Figure 2 、 Figure 3 、 Figure 4 First, install the collector 1 on the gluing robot and run the gluing program on the robot controller. After debugging the program of the gluing robot, the automatic teaching flow chart is as follows: Figure 3As shown, the specific implementation steps are as follows: the user runs the robot gluing program, the robot completes one gluing, and then the user manually checks the robot's gluing quality, selects ≥3 qualified gluing sections, and the user operates the gluing robot so that the gluing robot is located in the qualified gluing section, and the robot's posture is close to the working posture at that section. Through the interactive interface 13, the user sets the basic parameters 8, enters the work task number, part number, etc., and sets the items of visual inspection of gluing defects, such as size, etc. Through the interactive interface 13, the user clicks the one-key teaching 9 button, and the program automatically controls the collector to change the color of the acquisition fill light 4 and collect image photos at a fixed time. Like filtering, synthesis, binarization, comparison standards, feature recognition, calculation of detection item parameters, etc., the interactive interface 13 is displayed on the recognition display 16 to show the dynamic process of recognition. After the recognition display 16 task is completed, the final result will be displayed on the interactive interface 13 through the recognition display 16. The result will be distributed in the shape of the marking box 5. The user judges whether the marked range is correct. If it is correct, click the [Confirm] button. If it is not correct, click the [Cancel] button and re-recognize. After completing all teaching operations, press the [Finish] button to perform weighted evaluation on all teaching results. The sum of the weights is equal to 1. Finally, the task of visual inspection of glue defects is stored to complete.

[0058] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above and that the invention can be embodied in other specific forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments should be considered in all respects as illustrative and non-restrictive, and the scope of the invention is defined by the appended claims, not the foregoing description, and all variations within the meaning and range of equivalents of the claims are intended to be included therein. Any reference sign in a claim should not be construed as limiting the claim to which it relates.

Claims

1. An automatic teaching method for visual inspection of glue coating defects, characterized by: The system comprises a collector (1), a process module (7) and an interactive interface (13), and includes the following steps: Step 1: Install the collector (1) on the gluing robot and start the robot to complete one gluing operation; Step 2: Setting basic parameters (8) through the interactive interface (13) to set the project of visual inspection of glue defects, wherein the basic parameters (8) include the number of teaching times, the work task number, and the component number parameters, and the completion button can be used to calculate the weighted average value according to the results of each teaching time; Step 3: The user clicks on the one-button teaching (9) through the interactive interface (13), and the program automatically controls the collector (1) to change the color of the acquisition fill light (4) and collect image photos, and performs filtering, synthesis, binarization, comparison standards, feature recognition, calculation of detection project parameters, etc. on the image, and starts the automatic recognition task; Step 4: After the automatic recognition task is completed, the final result is displayed through the status display (17) on the interactive interface (13); Step 5: The user determines whether the marked range is correct. If it is correct, click the [Confirm] button on the interactive interface (13). If it is incorrect, click the [Cancel] button to re-identify. Step 6: Perform one-key teaching (9) for each selected qualified glue coating section in turn. After the teaching is completed, press the [Finish] key to perform weighted evaluation on all teaching results, and the sum of the weights is equal to 1. The glue coating defect visual inspection task is stored and the teaching is completed.

2. The automatic teaching method for visual inspection of glue coating defects according to claim 1, characterized in that: The collector (1) is composed of a glue gun (2), three collection cameras (3) and a collection fill light (4); the process module (7) includes setting basic parameters (8), one-key teaching (9), automatic recognition of user expectations (10), user confirmation (11) and task issuance (12); the collector (1) is electrically connected to the process module (7); and the interactive interface (13) is provided with a basic parameter area (14), one-key teaching (9), recognition display (16), status display (17) and a completion button.

3. The automatic teaching method for visual inspection of glue coating defects according to claim 2, characterized in that: The identification display (16) includes three acquisition cameras (3), a glue gun (2), a marking frame (5) and a glue strip (6).

4. The automatic teaching method for visual inspection of glue coating defects according to claim 2, characterized in that: The automatic identification of user expectations (10) is a program that automatically identifies user detection expectations based on the collected images. The automatic identification of user expectations (10) includes adjusting the colors of different fill lights, collecting images, image preprocessing, cutting pictures, identifying features, quantizing grayscale values ​​from 0 to 255, and calculating pixel values ​​of straight line widths.

5. The automatic teaching method for visual inspection of glue coating defects according to claim 1, characterized in that: The one-key teaching (9) user confirms the teaching preparation is complete through this command and starts the automatic recognition task.

6. The automatic teaching method for visual inspection of glue coating defects according to claim 2, characterized in that: The one-key teaching (9) includes displaying the identified user expectation on the interface, marking the rubber strip contour and edge, and the user confirmation (11) is to confirm that the marked rubber strip contour, edge and other information meet the user expectation.

7. The automatic teaching method for visual inspection of glue coating defects according to claim 2, characterized in that: The task delivery (12) includes saving the automatic identification of user expectations (10) and delivering related work software modules.

8. The automatic teaching method for visual inspection of glue coating defects according to claim 1, characterized in that: The collection fill light (4) is an RGB fill light.

Images