Generation method and system of visual inspection process file, and storage medium

Abstract

The invention discloses a visual inspection process file generation method and system and a storage medium. The visual inspection process file generation method comprises the following steps: acquiring a picture of a to-be-inspected product collected in a visual mode, and inputting the picture into a visual inspection software platform; according to a preset visual field, an artificial intelligence partitioning tool of a visual detection software platform is called to partition the picture of the to-be-detected product, and a to-be-detected region position frame group is generated; and calling a defect detection tool of the visual detection software platform, and generating a detection process file by taking the to-be-detected region position frame group as the input of the defect detection tool. Therefore, by adopting a defect detection tool which adopts lightweight AI deployment and integrates a detection algorithm formed by expert experience, on the basis of a visual software platform, rapid detection process file configuration is realized by automatically segmenting a to-be-detected area by AI and calling the defect detection tool in a partition manner, the establishment time of the detection process file is greatly shortened, and the detection efficiency is improved. And the production efficiency of field equipment is improved.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular, to a method and system for generating a visual inspection process document, and a storage medium. Background Art

[0002] In the field of visual inspection such as FPC (Flexible Printed Circuit), PCB (Printed Circuit Board), and AOI (Automated Optical Inspection), it is necessary to set inspection process documents according to different products. There are many disadvantages in this situation, including: a large variety; many positions to be inspected; a targeted imaging method required for each position; different standards need to be set for the separately imaged pictures according to the design; the actual imaging effect may deviate from the designed effect, and the standard setting of the inspection document needs to be iteratively adjusted according to the actual imaging effect and test results; for difficult-to-inspect pictures, multiple algorithms may need to be integrated; overall effect: a long time for new product introduction; a long debugging time to reach the mass production state.

[0003] The establishment of the existing inspection process documents is generally based on traditional vision methods. For traditional vision methods, there are multiple processing procedures involved. For example: there are many positions to be inspected, a targeted imaging method is required for each position, different standards need to be set for the separately imaged pictures according to the design, the actual imaging effect may deviate from the designed effect, the standard setting of the inspection document needs to be iteratively adjusted according to the actual imaging effect and test results, and for difficult-to-inspect pictures, multiple algorithms may need to be integrated. Establishing inspection process documents based on traditional vision methods will result in a long time for new product introduction and a long debugging time to reach the mass production state.

[0004] Therefore, the problems existing in the establishment of the existing inspection process documents based on traditional vision methods are: it takes too long to make visual inspection process documents for different products, resulting in a serious delay in the product production time (from several hours to several days), and it requires specialized engineers to perform daily maintenance operations, the process is cumbersome, time-consuming for adjustment, and it cannot meet the production requirements in a timely manner when there are many new products to be inspected. Summary of the Invention

[0005] Embodiments of the present invention aim to provide a method and system for generating a visual inspection process document, and a storage medium, aiming to solve the problems that it takes too long to make visual inspection process documents for different products in the existing situation, resulting in a serious delay in the product production time, and it requires specialized engineers to perform daily maintenance operations, the process is cumbersome, time-consuming for adjustment, and it cannot meet the production requirements in a timely manner when there are many new products to be inspected.

[0006] To solve the above technical problems, an embodiment of the first aspect of the present invention provides a method for generating a visual inspection process document, including:

[0007] Obtain the picture of the product to be inspected collected by visual means and input it into the visual inspection software platform;

[0008] Call the artificial intelligence partitioning tool of the visual inspection software platform according to the preset field of view to divide the picture of the product to be inspected, and generate a group of position frames for the areas to be inspected;

[0009] Call the defect detection tool of the visual inspection software platform, and use the group of position frames for the areas to be inspected as the input of the defect detection tool to generate a detection process document.

[0010] Optionally, the preset field of view includes a single field of view; the step of calling the artificial intelligence partitioning tool of the visual inspection software platform according to the preset field of view to divide the picture of the product to be inspected and generate a group of position frames for the areas to be inspected includes:

[0011] On the single field of view, call the template tool of the visual inspection software platform to frame the picture of the product to be inspected;

[0012] Call the artificial intelligence partitioning tool of the visual inspection software platform to frame and divide the picture of the product to be inspected in the template tool into each area to be inspected. The artificial intelligence algorithm automatically identifies each area to be inspected in the picture of the product to be inspected, generates a rectangular frame including each area to be inspected, and the rectangular frames of each area to be inspected form a group of position frames for the areas to be inspected.

[0013] Optionally, the preset field of view includes a stitched field of view; the step of calling the artificial intelligence partitioning tool of the visual inspection software platform according to the preset field of view to divide the picture of the product to be inspected and generate a group of position frames for the areas to be inspected includes:

[0014] On the stitched field of view, use the central area of 4 / 5 of the picture of the product to be inspected as the template tool;

[0015] Use the selected template tool above to frame the picture of the product to be inspected;

[0016] Call the artificial intelligence partitioning tool of the visual inspection software platform to frame and divide the picture of the product to be inspected in the selected template tool above into each area to be inspected. The artificial intelligence algorithm automatically identifies each area to be inspected in the picture of the product to be inspected, generates a rectangular frame including each area to be inspected, and the rectangular frames of each area to be inspected form a group of position frames for the areas to be inspected.

[0017] Optionally, the step of calling the defect detection tool of the visual inspection software platform and using the group of position frames for the areas to be inspected as the input of the defect detection tool to generate a detection process document includes:

[0018] Invoke the defect detection tool of the visual detection software platform;

[0019] Use the rectangular frames of the areas to be inspected as the input of the defect detection tool, and convert the positions of the rectangular frames of the areas to be inspected into a number of defect tools using the default parameters of the defect detection tool, generating a detection process file including the number of defect tools.

[0020] Optionally, the method for generating the visual detection process file further includes: according to the detection standard file of the product, checking and filling the generated detection process file to establish a final detection process file.

[0021] Optionally, the step of checking and filling the generated detection process file according to the detection standard file of the product to establish a final detection process file includes:

[0022] According to the detection standard file of the product, in the generated detection process file, confirm the areas to be inspected that are not recognized by the artificial intelligence partitioning tool, and add the unrecognized areas to be inspected to the detection process file by box selection;

[0023] Confirm the areas to be inspected that do not need to be detected generated by the artificial intelligence partitioning tool, and delete the areas to be inspected that do not need to be detected;

[0024] Use the detection process file after the above confirmation operation as the final detection process file.

[0025] Optionally, the method for generating the visual detection process file further includes: when the detection standard changes, make a partial adjustment to the invoked defect detection tool, and fine-tune the defect tool parameters in the visual detection process file to form a new visual detection process file.

[0026] Correspondingly, an embodiment of the second aspect of the present invention provides a system for generating a visual detection process file, including: a vision device and an image processing device, where:

[0027] The vision device collects pictures of the product to be inspected visually and transmits them to the image processing device through a preset transmission method;

[0028] The image processing device is built with a visual detection software platform, divides the pictures of the product to be inspected by invoking the artificial intelligence partitioning tool of the visual detection software platform to generate a set of position frames of the areas to be inspected, and invokes the defect detection tool of the visual detection software platform, using the set of position frames of the areas to be inspected as the input of the defect detection tool to generate a detection process file.

[0029] Accordingly, an embodiment of the third aspect of the present invention provides a system for generating a visual inspection process file, including an image processing device. The image processing device includes a memory, a processor, and a computer program stored on the memory and running on the processor. When the computer program is executed by the processor, it implements the method for generating a visual inspection process file according to the embodiment of the first aspect of the present invention.

[0030] Accordingly, an embodiment of the fourth aspect of the present invention provides a storage medium with a program for the method of generating a visual inspection process file. When the program for the method of generating a visual inspection process file is executed by a processor, it implements the method for generating a visual inspection process file according to the embodiment of the first aspect of the present invention.

[0031] Compared with the prior art, an embodiment of the present invention provides a method, a system, and a storage medium for generating a visual inspection process file. The method for generating a visual inspection process file includes: obtaining a picture of a product to be inspected collected visually and inputting it into a visual inspection software platform; dividing the picture of the product to be inspected by calling an artificial intelligence partitioning tool of the visual inspection software platform according to a preset field of view to generate a group of position frames for the area to be inspected; calling a defect detection tool of the visual inspection software platform, using the group of position frames for the area to be inspected as the input of the defect detection tool, and generating a detection process file. Thus, by adopting lightweight AI deployment and a defect detection tool integrating detection algorithms formed by expert experience, based on a visual software platform, automatically dividing the area to be inspected by AI and calling the defect detection tool by partition to achieve rapid configuration of the detection process file, significantly shortening the establishment time of the detection process file and improving the production efficiency of on-site equipment. Thereby, it can solve the problems that the existing method of making visual inspection process files for different products takes too long, resulting in serious delays in product production time, and requires specialized engineers for daily maintenance operations, with a cumbersome process, time-consuming adjustment, and inability to meet production needs in a timely manner when there are many new products to be inspected. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] One or more embodiments are exemplarily illustrated by pictures in the corresponding drawings. These exemplary illustrations do not limit the embodiments. Elements with the same reference numerals in the drawings represent similar elements. Unless otherwise stated, the drawings in the figures do not constitute a scale limitation.

[0033] Figure 1 is a schematic structural diagram of a system for generating a visual inspection process file provided by the present invention;

[0034] Figure 2 is a schematic flowchart of a method for generating a visual inspection process file provided by the present invention;

[0035] Figure 3It is a schematic diagram of dividing the image of a product to be inspected into multiple appropriately sized rectangular ROI regions in the AI algorithm of a method for generating a visual inspection process document provided by the present invention;

[0036] Figure 4 It is another process schematic diagram of a method for generating a visual inspection process document provided by the present invention;

[0037] Figure 5 It is a process schematic diagram of applying a method for generating a visual inspection process document provided by the present invention to a connector PCB;

[0038] Figure 6 It is a structural schematic diagram of an image processing device in a system for generating a visual inspection process document provided by the present invention. Detailed implementation manners

[0039] For the convenience of understanding the present invention, the present invention will be described in more detail below with reference to the accompanying drawings and specific embodiments. It should be noted that when an element is expressed as "fixed to" another element, it can be directly on the other element, or there can be one or more intermediate elements therebetween. When an element is expressed as "electrically connected to" another element, it can be directly connected to the other element, or there can be one or more intermediate elements therebetween. The terms "upper", "lower", "inner", "outer", "bottom", etc. used in this specification indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus cannot be understood as a limitation to the present invention. In addition, the terms "first", "second", "third", etc. are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0040] Unless otherwise defined, all technical and scientific terms used in this specification have the same meaning as commonly understood by those skilled in the technical field to which the present invention belongs. The terms used in the specification of the present invention are only for the purpose of describing specific embodiments and are not used to limit the present invention. The term "and / or" used in this specification includes any and all combinations of one or more of the related listed items.

[0041] In addition, the technical features involved in different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0042] In one embodiment, as Figure 1 shown, the present invention provides a system 10 for generating a visual inspection process document, including: a vision device 100 and an image processing device 900, wherein:

[0043] The vision device 100, including at least one camera, captures images of the product to be inspected visually and transmits them to the image processing device 900 through a preset transmission method.

[0044] The image processing device 900 has a vision detection software platform built-in. It divides the images of the product to be inspected by invoking the artificial intelligence (AI) zoning tool of the vision detection software platform, generates a set of position frames for the areas to be inspected, and invokes the defect detection tool of the vision detection software platform. Using the set of position frames for the areas to be inspected as the input of the defect detection tool, it generates a detection process file. For example, the image processing system can be a computer.

[0045] The image processing device 900 includes a display device for displaying the result output. The display device can be a monitor.

[0046] The image processing device 900 includes an input device for human-computer interaction. The input device can be a keyboard.

[0047] The image processing device 900 includes a graphical input device for graphical human-computer interaction. The graphical input device can be a mouse.

[0048] In this embodiment, by providing a system for generating a vision detection process file, including a vision device and an image processing device, wherein: the vision device captures images of the product to be inspected visually and transmits them to the image processing device through a preset transmission method; the image processing device has a vision detection software platform built-in. It divides the images of the product to be inspected by invoking the artificial intelligence zoning tool of the vision detection software platform, generates a set of position frames for the areas to be inspected, and invokes the defect detection tool of the vision detection software platform. Using the set of position frames for the areas to be inspected as the input of the defect detection tool, it generates a detection process file. Thus, by adopting lightweight AI deployment and a defect detection tool integrating expert experience-formed detection algorithms, based on the vision software platform, combined with the product characteristics of vision detection, using AI to automatically segment the areas to be inspected and call the defect detection tool for each area to achieve rapid configuration of the detection process file, significantly shortening the establishment time of the detection process file (through actual testing, the establishment time of the detection process file can be shortened by more than 90%), and improving the production efficiency of on-site equipment. Thereby, it can solve the problems that the existing method of creating vision detection process files for different products takes too long, resulting in a serious delay in the product production time, and requires specialized engineers for daily maintenance operations, with a cumbersome process, time-consuming adjustment, and inability to meet the production demand in a timely manner when there are many new products to be inspected.

[0049] In one embodiment, as Figure 2 shown, the present invention provides a method for generating a vision detection process file, including:

[0050] S1. Obtain the picture of the product to be inspected collected by visual means and input it into the visual inspection software platform;

[0051] S2. Call the artificial intelligence (AI) zoning tool of the visual inspection software platform according to the preset field of view to divide the picture of the product to be inspected, and generate a group of position frames for the area to be inspected;

[0052] S3. Call the defect detection tool of the visual inspection software platform, use the group of position frames for the area to be inspected as the input of the defect detection tool, and generate a detection process file.

[0053] In this embodiment, by providing a method for generating a visual inspection process file, including: obtaining the picture of the product to be inspected collected by visual means and inputting it into the visual inspection software platform; calling the AI zoning tool of the visual inspection software platform according to the preset field of view to divide the picture of the product to be inspected, and generating a group of position frames for the area to be inspected; calling the defect detection tool of the visual inspection software platform, using the group of position frames for the area to be inspected as the input of the defect detection tool, and generating a detection process file. Thus, by adopting a lightweight AI deployment and a defect detection tool integrated with an inspection algorithm formed by expert experience, based on the visual software platform, combined with the product characteristics of visual inspection, the area to be inspected is automatically segmented by AI and the defect detection tool is called in a partitioned manner to achieve a rapid configuration of the inspection process file, greatly shortening the establishment time of the inspection process file (through actual testing, the establishment time of the inspection process file can be shortened by more than 90%), and improving the production efficiency of on-site equipment. Thereby, it can solve the problems that the existing method of making visual inspection process files for different products takes too long, resulting in a serious delay in the product production time, and requires specialized engineers to perform daily maintenance operations, the process is cumbersome, the adjustment is time-consuming, and it cannot meet the production requirements in a timely manner when there are many new products to be inspected.

[0054] In one embodiment, in step S1, obtain the picture of the product to be inspected collected by visual means and input it into the visual inspection software platform.

[0055] The visual inspection software platform is a multi-camera visual inspection software platform, which is applied to usage scenarios that require multi-camera composite inspection such as optical sorting machines and die-cutting inspection machines, and is mainly used to realize various inspection requirements such as the size, defects, dirt, wrinkles, characters, two-dimensional codes, etc. of materials, and supports AI artificial intelligence inspection. The visual inspection software platform includes a template tool, an AI zoning tool, and a defect detection tool, where:

[0056] The AI zoning tool can divide the picture of the product to be inspected and generate a group of position frames for the area to be inspected.

[0057] The template tool is a tool in the visual inspection software platform that establishes the detection reference for the product to be inspected by selecting the target product to be inspected by framing. The specific functions implemented by the template tool are: selecting the picture of the target product to be inspected as the target reference that the visual inspection software platform searches for in the image during the subsequent inspection process. The detection positions of all subsequent detection tools are determined based on the relative positions located by the matched template tool.

[0058] The defect detection tool integrates the detection algorithm formed by expert experience, which is used to take the rectangular frames R1 - Rm as the input of the defect detection tool, and uses the default parameters of the defect detection tool to convert the positions of all rectangular frames R1 - Rm into the defect detection tools T1 - Tm, generating a detection process file including defective tools.

[0059] The position of the detection area is the relative position with respect to the positioning of the template tool. The position of one template tool determines the relative positions of a group of detection tools associated with this template tool. This ensures that when the product appears in the field of view, whether it is offset or rotated, as long as the position and angle after the template tool is matched are determined, the detection area will appear at the correct position of the product to be inspected. That is, various detection tools are established under the product coordinate system, rather than the world coordinate system.

[0060] In the present invention, the relationships between the coordinate systems of each component are as follows:

[0061] Picture: Based on the picture coordinate system.

[0062] Template: Through template matching, determine the position and angle of a single product coordinate system in the picture coordinate system.

[0063] Detection area: Determine its own position in the product coordinate system, generally a rectangular frame, which can be understood as a local small picture cut out from the entire product picture.

[0064] Detection tool: The input of all detection tools is a rectangular picture. The detection area, as a small rectangular picture, can be used as the input and called by the corresponding detection tool, and then the corresponding algorithm can be used to process and output the detection result; in the present invention, the main detection tool used is the defect tool. After the defect tool batch reads the detection areas generated by AI, a batch of defect tools are generated.

[0065] The picture of the product to be inspected can be obtained by a vision device through vision. For example, the vision device consists of multiple cameras. The vision device composed of multiple cameras collects the picture of the product to be inspected through vision and transmits it to the image processing device with the visual inspection software platform built in through a preset transmission method. The collection method or collection method for the vision device to collect the picture of the product to be inspected through vision can be implemented by existing vision image collection methods, which will not be elaborated here.

[0066] In one embodiment, in step S2, according to a preset field of view, the AI partitioning tool of the visual detection software platform is called to divide the picture of the product to be inspected, generating a set of position frames for the areas to be inspected.

[0067] Specifically, the preset field of view includes a single field of view and a stitched field of view. The single field of view can cover one or more products to be inspected within one field of view, and is applicable to scenarios where there is one or more products to be inspected within one field of view. The stitched field of view can only cover 1 / n of the product to be inspected, and n fields of view need to be stitched to form a complete field of view map to cover one product to be inspected, and is applicable to scenarios of a large product to be inspected.

[0068] In step S2, according to the preset field of view, the AI partitioning tool of the visual detection software platform is called to divide the picture of the product to be inspected, generating a set of position frames for the areas to be inspected, including:

[0069] For the single field of view, the template tool of the visual detection software platform is called, and the AI partitioning tool of the visual detection software platform is called to circle and divide the picture of the product to be inspected, generating a set of position frames for the areas to be inspected; specifically including:

[0070] On the single field of view, the template tool of the visual detection software platform is called to frame the picture of the product to be inspected;

[0071] The AI partitioning tool of the visual detection software platform is called to frame and divide the picture of the product to be inspected in the template tool into each area to be inspected. The AI algorithm automatically identifies each area to be inspected in the picture of the product to be inspected, generating rectangular frames R1 - Rm including each area to be inspected. The rectangular frames R1 - Rm of each area to be inspected form a set of position frames for the areas to be inspected. Thus, the relationship of "template shape and position - one - to - one associated with detection area shape and position" is established in the detection process document.

[0072] Among them, the AI algorithm, according to the color distribution of the overall image in the picture of the product to be inspected, automatically divides areas with similar colors or areas judged as a whole device (such as resistors, capacitors, inductors, heat sinks, etc.) into multiple rectangular ROI (Region of Interest) areas of appropriate sizes. These rectangular ROI areas are the rectangular frames R1 - Rm of each area to be inspected (for example, the rectangular frames 1, 2, 3, n shown by the dotted lines in Figure 3 ). After generating the rectangular ROI areas, the relevant rectangular ROI areas are used as the input of the defect detection tool, and an algorithm is called for subsequent defect detection. That is: the AI algorithm only makes the division of the rectangular frames R1 - Rm of each area to be inspected, and the defect detection tool performs detection operations separately according to each divided area to be inspected.

[0073] Before using the AI algorithm to automatically identify the areas to be inspected in the image of the product to be inspected, the electronic components are pre-trained using AI to identify the area including the electronic components as a rectangular ROI area, forming a pre-trained device recognition model. The pre-trained device recognition model is then called in implementation to divide the device as a whole and identify the rectangular ROI area including the electronic components.

[0074] For the stitching field of view, on the entire stitching field of view, the central area of 4 / 5 of the image of the product to be inspected is automatically used as a template tool, and the AI partitioning tool of the visual inspection software platform is called to circle the image of the product to be inspected for division, and generate a position frame group of the area to be inspected; specifically, it includes:

[0075] In the stitching field of view, the central area of 4 / 5 of the image of the product to be inspected is used as a template tool.

[0076] When the field of view is smaller than the product to be inspected, the image captured by the camera is associated with the movement (xn, yn); due to product positioning, the product position corresponding to the movement (xn, yn) coordinates when the camera takes the picture is certain, and the product's own shape is consistent, so the captured partial image of the product P (xn, yn) is also basically consistent.

[0077] In the actual environment, even if the same position is photographed twice, the imaging results will have slight changes, and the image effects may be slightly offset due to tolerance between products, so it is necessary to do a second precision positioning of the area to be inspected for accurate inspection. Therefore, for the picture taken at the (xn, yn) position, the software automatically selects the image of the center area as the template tool for secondary positioning. Preferably, the center area of 4 / 5 of the picture of the product to be inspected is automatically selected as the template tool for secondary positioning.

[0078] Use the template tool selected above to frame the image of the product to be inspected.

[0079] The AI partitioning tool of the visual inspection software platform is called to select the image of the product to be inspected in the template tool selected above and divide it into various inspection areas. The AI algorithm automatically identifies the various inspection areas in the image of the product to be inspected and generates rectangular frames R1-Rm including the various inspection areas. The rectangular frames R1-Rm of the various inspection areas form the inspection area position frame group. At this point, the "template shape and position are associated with the inspection area shape and position" relationship is established in the inspection process file.

[0080] In this embodiment, by calling the AI partitioning tool of the visual inspection software platform according to the preset field of view to partition the picture of the product to be inspected, a set of position frames of the areas to be inspected is generated. Thus, by adopting lightweight AI deployment, the areas to be inspected and the partitions can be automatically segmented by AI, and the relationship of "template shape and position - associated detection area shape and position" is established in the inspection process file, providing accurate input conditions for subsequent defect detection tools.

[0081] In one embodiment, in step S3, the defect detection tool of the visual inspection software platform is called, and the set of position frames of the areas to be inspected is used as the input of the defect detection tool to generate an inspection process file; specifically including:

[0082] S31. Call the defect detection tool of the visual inspection software platform;

[0083] S32. Take the rectangular frames R1 - Rm of each area to be inspected as the input of the defect detection tool, and use the default parameters of the defect detection tool to convert the positions of the rectangular frames R1 - Rm of each area to be inspected into a number of defect tools T1 - Tm, generating an inspection process file including a number of defect tools.

[0084] So far, the relationship of "template shape and position - associated detection area shape and position - associated defect detection tool and parameter configuration" is established in the inspection process file.

[0085] The defect detection tool of the visual inspection software platform is used to detect the defects of the product to be tested. The defect detection tool includes parameters such as tolerance setting, defect parameters, pixel difference amount, proportion difference amount, and proportion pixel value, where:

[0086] Tolerance setting: The standard value of the defect can be set. If the difference between the corresponding area of the image during operation and the template is less than the currently set standard value, the defect detection passes and is represented in green; otherwise, it indicates that the defect detection fails and is represented in red.

[0087] Defect parameters: Used to adjust the effect of the defect detection tool, including five basic parameters: boundary recognition threshold, noise filtering size, boundary width, contrast within the boundary range, and contrast outside the boundary range. Among them, the specific descriptions of each parameter are as follows:

[0088] Boundary recognition threshold: This parameter is mainly used to set the contrast of the boundary of the area to be inspected. The smaller the value, the more sensitive the boundary contrast. Its range is [5 - 150], and the default value of this parameter is 10. If many small green interference points appear within the red square when using the default value, this value can be appropriately increased to reduce interference. In the actual use process, users need to reasonably set this value according to the specific product image.

[0089] Filter noise size: This parameter is used to set the size of the noise area to be filtered. The larger the value, the larger the noise area to be filtered. Conversely, the smaller the noise area to be filtered.

[0090] Boundary width: This parameter represents the boundary tolerance during threshold segmentation. The valid range of its value is [0.5, 40]. The larger the value, the higher the tolerance. Conversely, the smaller the tolerance. The default value of this parameter is 10. Unless there are special circumstances, the default value is generally used.

[0091] Contrast within the boundary range: This parameter represents the contrast of the product within the recognized contour (green contour). The valid value range is [0 - 255]. This parameter generally uses the default value of 30. Users can also set a reasonable value according to the image contrast.

[0092] Contrast outside the boundary range: This parameter represents the contrast of the product outside the recognized contour (green contour). The valid value range is [0 - 255]. This parameter generally uses the default value of 30. Users can also set a reasonable value according to the image contrast.

[0093] Difference in pixel quantity: This parameter represents the pixel difference between the current defect area to be detected and the defect template area. If the pixel difference is greater than the set difference value, the defect detection of the area to be inspected fails and is marked in red. Otherwise, it passes and is marked in green.

[0094] Ratio of difference in pixel quantity: Set the proportion of the difference in pixel quantity, default is 0.05.

[0095] Proportional pixel value: The size of the tool ROI area multiplied by the ratio of the difference in pixel quantity.

[0096] In the visual inspection software platform, left - click on the defect detection tool with the mouse. Then left - click on the mouse in the middle position on the left side of the area to be inspected. At this time, a gray dot will appear. Then move the mouse. During the mouse movement, a gray straight line will be automatically generated. After the mouse moves to the middle position on the other side of the area to be inspected, left - click on the mouse again. Then move the mouse up and down, and a gray rectangular box will appear. After the rectangular box completely encloses the area to be inspected, left - click on the mouse again to generate a red square box.

[0097] In this embodiment, by invoking the defect detection tool of the vision detection software platform and using the position frame group of the area to be inspected as the input of the defect detection tool, a detection process file is generated. Thus, by adopting a lightweight AI deployment and a defect detection tool with a detection algorithm integrating expert experience, based on the vision software platform, combined with the product characteristics of vision detection, the AI automatically segments the area to be inspected and calls the defect detection tool in partitions to achieve a rapid configuration of the detection process file, significantly shortening the establishment time of the detection process file (through actual tests, the establishment time of the detection process file can be shortened by more than 90%), and improving the production efficiency of on-site equipment.

[0098] In one embodiment, as Figure 4 shown, the method for generating the vision detection process file further includes: S4. According to the detection standard file of the product, check for omissions and fill in the blanks in the generated detection process file to establish the final detection process file; specifically including:

[0099] S41. According to the detection standard file of the product, in the above-generated detection process file, confirm the areas to be inspected that have not been recognized by the AI partitioning tool, and add the unrecognized areas to the detection process file by framing;

[0100] S42. Confirm the areas to be inspected that do not need to be detected generated by the AI partitioning tool, and delete these areas that do not need to be detected;

[0101] S43. Use the detection process file after the operations in steps S41 and S42 as the final detection process file.

[0102] In this embodiment, by checking for omissions and filling in the blanks in the generated detection process file according to the detection standard file of the product, it can be confirmed which areas to be inspected have not been recognized by the AI partitioning tool, so that they can be added to the detection process file by framing through other means (such as manual means), thus avoiding omission of the areas to be inspected; it can also be confirmed that the areas to be inspected generated by the AI partitioning tool actually do not need to be detected, and they can be deleted from the generated detection process file through other means (such as manual means) to reduce the subsequent detection time and improve the subsequent detection efficiency. Thus, by checking for omissions and filling in the blanks in the generated detection process file in combination with the detection standard file of the product, the finally generated detection process file can better meet the actual detection requirements, and can achieve a rapid configuration of the detection process file, significantly shortening the establishment time of the detection process file (through actual tests, the establishment time of the detection process file can be shortened by more than 90%), and improving the production efficiency of on-site equipment.

[0103] In one embodiment, as Figure 4As shown, the method for generating the visual inspection process document further includes: S5. When the inspection standard changes, locally adjust the defect detection tool called, and fine-tune the defect tool parameters in the visual inspection process document to form a new visual inspection process document.

[0104] Specifically, the local adjustment refers to operations such as adjusting the size, changing the position, and adjusting the detection parameters of a single defect tool.

[0105] When the inspection standard changes, by locally adjusting the defect detection tool called, fine-tuning the defect tool parameters in the visual inspection process document to form a new visual inspection process document, it is possible to quickly configure the inspection process document by combining the fine-tuning of standard parameters, significantly shortening the establishment time of the inspection process document (through actual testing, the establishment time of the inspection process document can be shortened by more than 90%), and improving the production efficiency of on-site equipment.

[0106] To facilitate the understanding of the above inventive concept of the present invention, the following will further elaborate on the above inventive concept of the present invention in conjunction with the accompanying drawings and specific embodiments.

[0107] As Figure 5 shown, an embodiment of the present invention provides a method for generating a visual inspection process document, which is applied to the method for generating a visual inspection process document for the wrong, missing, and reverse detection (wrong parts, missing installation, reverse installation) of a certain connector PCB, including:

[0108] S601. Obtain the color picture of the PCB product and input it into the visual inspection software platform.

[0109] S602. Establish a template tool for the color picture of the PCB product as the basis for subsequent positioning.

[0110] S603. Call the AI partitioning tool of the visual inspection software platform to partition the color picture of the PCB product and generate detection frames for each position of a single plug-in workpiece.

[0111] S604. Call the defect detection tool of the visual inspection software platform, use the detection frames of each position as the input of the defect detection tool, automatically generate a defect tool list for the detection frames of each position, and configure the parameters of the defect tool with preset default parameters to generate a detection process document including several defect tools.

[0112] S605. According to the product inspection standard document, check and fill in the deficiencies of the generated detection process document to establish the final detection process document.

[0113] S606. When the inspection standard changes, locally adjust the defect detection tool called, and fine-tune the defect tool parameters in the visual inspection process document to form a new visual inspection process document.

[0114] S607. Save the visual inspection process file and use the visual inspection process file for the production of PCB products.

[0115] In this embodiment, by adopting a lightweight AI deployment and a defect detection tool integrated with a detection algorithm formed by expert experience, based on a visual software platform, combined with the characteristics of PCB products for visual inspection, the AI is used to automatically segment the area to be inspected and partitionally call the defect detection tool to achieve a rapid configuration of the inspection process file, significantly shortening the establishment time of the inspection process file (through actual testing, the establishment time of the inspection process file can be reduced from the original 120 minutes to 5 minutes, shortening 95% of the establishment time of the inspection process file), and improving the production efficiency of on-site equipment.

[0116] Based on the same concept, in one embodiment, the present invention further provides a visual inspection process file generation system 10, including an image processing device 900, as Figure 6 shown. The image processing device 900 includes: a memory 902, a processor 901, and one or more computer programs stored in the memory 902 and executable on the processor 901. The memory 902 and the processor 901 are coupled together through a bus system 903. When the one or more computer programs are executed by the processor 901, the following steps of a visual inspection process file generation method provided by an embodiment of the present invention are implemented:

[0117] S1. Obtain the picture of the product to be inspected collected by visual means and input it into the visual inspection software platform;

[0118] S2. Call the artificial intelligence partitioning tool of the visual inspection software platform according to a preset field of view to divide the picture of the product to be inspected, and generate a group of position frames for the area to be inspected;

[0119] S3. Call the defect detection tool of the visual inspection software platform, use the group of position frames for the area to be inspected as the input of the defect detection tool, and generate an inspection process file.

[0120] The method disclosed in the embodiments of the present invention described above may be applied to or implemented by the processor 901. The processor 901 may be an integrated circuit chip with signal processing capabilities. During implementation, the steps of the above method may be completed by the integrated logic circuit in hardware or instructions in software form in the processor 901. The processor 901 may be a general-purpose processor, a DSP (Digital Signal Processor), an MCU (Micro Control Unit), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The processor 901 may implement or execute the various methods, steps, and logic block diagrams disclosed in the embodiments of the present invention. The general-purpose processor may be a microprocessor or any conventional processor, etc. Combining the steps of the method disclosed in the embodiments of the present invention, it may be directly embodied as being executed and completed by the hardware decoding processor, or executed and completed by the combination of the hardware and software modules in the decoding processor. The software module may be located in the storage medium, and this storage medium is located in the memory 902. The processor 901 reads the information in the memory 902 and combines its hardware to complete the steps of the foregoing method.

[0121] It can be understood that the memory 902 in the embodiments of the present invention can be a volatile memory or a non-volatile memory, or can include both volatile and non-volatile memories. Among them, the non-volatile memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable read-only memory (EEPROM), a ferromagnetic random access memory (FRAM), a flash memory or other memory technologies, a compact disk read-only memory (CD-ROM), a digital versatile disk (DVD) or other optical disk storage, a magnetic cassette, a magnetic tape, a magnetic disk storage or other magnetic storage devices; the volatile memory can be a random access memory (RAM). By way of example but not limitation, many forms of RAM are available, such as a static random access memory (SRAM), a synchronous static random access memory (SSRAM), a dynamic random access memory (DRAM), a synchronous dynamic random access memory (SDRAM), a double data rate synchronous dynamic random access memory (DDR SDRAM), an enhanced synchronous dynamic random access memory (ESDRAM), a sync link dynamic random access memory (SLDRAM), a direct rambus random access memory (DRRAM).The memories described in the embodiments of the present invention are intended to include, but are not limited to, these and any other suitable types of memories.

[0122] It should be noted that the above embodiments of the laminating and gluing whole-line equipment and the method embodiments belong to the same concept. For the specific implementation process, please refer to the method embodiments. And the technical features in the method embodiments are all correspondingly applicable to the embodiments of the laminating and gluing whole-line equipment, which will not be elaborated here.

[0123] In addition, in an exemplary embodiment, the embodiments of the present invention also provide a computer storage medium, specifically a computer-readable storage medium. For example, it includes a memory 902 storing a computer program. One or more programs of a method for generating a vision inspection process file are stored on the computer storage medium. When the one or more programs of the method for generating a vision inspection process file are executed by a processor 901, the following steps of a method for generating a vision inspection process file provided by the embodiments of the present invention are implemented:

[0124] S1. Obtain a picture of a product to be inspected collected by a vision method and input it into a vision inspection software platform;

[0125] S2. Call the artificial intelligence partitioning tool of the vision inspection software platform according to a preset field of view to divide the picture of the product to be inspected, and generate a group of position frames for the area to be inspected;

[0126] S3. Call the defect detection tool of the vision inspection software platform, and use the group of position frames for the area to be inspected as the input of the defect detection tool to generate a detection process file.

[0127] It should be noted that the program embodiments of the method for generating a vision inspection process file on the above computer-readable storage medium and the method embodiments belong to the same concept. For the specific implementation process, please refer to the method embodiments. And the technical features in the method embodiments are all correspondingly applicable to the embodiments of the above computer-readable storage medium, which will not be elaborated here.

[0128] It should be noted that in this article, the term "including", "comprising" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or further includes elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "including one..." does not exclude the existence of another identical element in the process, method, article or device including that element.

[0129] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; under the idea of the present invention, the technical features in the above embodiments or different embodiments can also be combined, and the steps can be implemented in any order, and there are many other variations in different aspects of the present invention as described above. For the sake of brevity, they are not provided in detail; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for generating a visual inspection process document, characterized in that Including: Obtain the picture of the product to be inspected collected visually and input it into the visual inspection software platform; Call the artificial intelligence zoning tool of the visual inspection software platform according to the preset field of view to divide the picture of the product to be inspected, and generate a group of position frames for the areas to be inspected; Call the defect detection tool of the visual inspection software platform, use the group of position frames for the areas to be inspected as the input of the defect detection tool, and generate a detection process file.

2. The method for generating a visual inspection process document according to claim 1, wherein The preset field of view includes a single field of view; the step of calling the artificial intelligence zoning tool of the visual inspection software platform according to the preset field of view to divide the picture of the product to be inspected and generate a group of position frames for the areas to be inspected includes: On the single field of view, call the template tool of the visual inspection software platform to frame the picture of the product to be inspected; Call the artificial intelligence zoning tool of the visual inspection software platform to frame and divide the picture of the product to be inspected in the template tool into each area to be inspected, the artificial intelligence algorithm automatically identifies each area to be inspected in the picture of the product to be inspected, generates a rectangular frame including each area to be inspected, and the rectangular frames of each area to be inspected form a group of position frames for the areas to be inspected.

3. The method for generating a visual inspection process document according to claim 1, wherein The preset field of view includes a stitched field of view; the step of calling the artificial intelligence zoning tool of the visual inspection software platform according to the preset field of view to divide the picture of the product to be inspected and generate a group of position frames for the areas to be inspected includes: On the stitched field of view, use the central area of 4 / 5 of the picture of the product to be inspected as the template tool; Use the above-selected template tool to frame the picture of the product to be inspected; Call the artificial intelligence zoning tool of the visual inspection software platform to frame and divide the picture of the product to be inspected in the above-selected template tool into each area to be inspected, the artificial intelligence algorithm automatically identifies each area to be inspected in the picture of the product to be inspected, generates a rectangular frame including each area to be inspected, and the rectangular frames of each area to be inspected form a group of position frames for the areas to be inspected.

4. The method for generating a visual inspection process document according to claim 1, wherein, The step of calling the defect detection tool of the visual inspection software platform, using the group of position frames for the areas to be inspected as the input of the defect detection tool, and generating a detection process file includes: Call the defect detection tool of the visual inspection software platform; Use the rectangular frames of each area to be inspected as the input of the defect detection tool, and use the default parameters of the defect detection tool to convert the positions of all the rectangular frames of each area to be inspected into several defect tools, and generate a detection process file including the several defect tools.

5. The method for generating a visual inspection process document according to claim 1, wherein The method for generating the visual inspection process file further includes: according to the product detection standard file, check for omissions and make up for deficiencies in the generated detection process file, and establish the final detection process file.

6. The method for generating a visual inspection process document according to claim 5, wherein The step of according to the product detection standard file, checking for omissions and making up for deficiencies in the generated detection process file, and establishing the final detection process file includes: According to the product detection standard file, in the generated detection process file, confirm the areas to be inspected that have not been identified by the artificial intelligence zoning tool, and frame and add the unrecognized areas to be inspected to the detection process file; Confirm the areas to be inspected that do not need to be detected generated by the artificial intelligence zoning tool, and delete the areas to be inspected that do not need to be detected; Use the inspection process document after the above confirmation operation as the final inspection process document.

7. The method for generating a visual inspection process document according to claim 1, wherein The method for generating the visual inspection process document further includes: when the inspection standard changes, locally adjust the called defect detection tool, and fine-tune the defect tool parameters in the visual inspection process document to form a new visual inspection process document.

8. A system for generating a visual inspection process document, characterized in that, Including: A vision device and an image processing device, where: The vision device collects pictures of the product to be inspected visually and transmits them to the image processing device through a preset transmission method; The image processing device is built-in with a visual inspection software platform, divides the pictures of the product to be inspected by calling the artificial intelligence partitioning tool of the visual inspection software platform to generate a group of position frames for the area to be inspected, and calls the defect detection tool of the visual inspection software platform. Using the group of position frames for the area to be inspected as the input of the defect detection tool, a detection process document is generated.

9. A system for generating a visual inspection process document, characterized in that, Including an image processing device, the image processing device includes a memory, a processor, and a computer program stored on the memory and running on the processor. When the computer program is executed by the processor, it implements the method for generating a visual inspection process document according to any one of claims 1 to 7.

10. A storage medium, characterized in that, A program for the method for generating a visual inspection process document is stored on the storage medium. When the program for the method for generating a visual inspection process document is executed by a processor, it implements the method for generating a visual inspection process document according to any one of claims 1 to 7.

Images