Visual detection method for circuit board defects, electronic equipment and storage medium

Through the positioning, classification and detection module of the intelligent visual recognition platform, combined with the point data of AOI equipment, the missed judgment and wrong judgment problems in complex defect detection are solved, and the precise positioning and identification of circuit board defects is realized, and the detection is improved. Accuracy and efficiency.

CN118941499AInactive Publication Date: 2024-11-12YITONG TECH (SHENZHEN) CO LTD

Patent Information

Application Number
CN202410877497.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-02
Publication Date
2024-11-12
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

The existing AI visual detection methods are prone to missed or misjudged when detecting complex defects, and cannot accurately determine the location and size of the defects, resulting in low detection accuracy.

Method used

Using an intelligent visual recognition platform, combining positioning modules, classification modules and detection modules, AOI devices obtain point data, determine the area to be detected, identify the device type and defect category, extract feature vectors, locate and output the location and size of the defect area .

Benefits of technology

It significantly improves the defect detection rate and detection accuracy, reduces false alarms and missed detection, and improves detection efficiency and accuracy.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118941499A_ABST
    Figure CN118941499A_ABST
Patent Text Reader

Abstract

The invention discloses a visual detection method for circuit board defects, electronic equipment and a storage medium, and the method comprises the steps: obtaining report point data through AOI equipment, determining a to-be-detected target region on a defect image through a positioning module, recognizing the type of a to-be-detected object contained in the target region through a classification module, and carrying out the recognition of the type of the to-be-detected object. Positioning a defect area with defects in the target area through a detection module according to the category of the to-be-detected object and the defect category of the defect information, and outputting a detection result according to the position and the size of the defect area; according to the embodiment of the invention, the defect position is determined by reducing the detection range, and the defect detection rate and the detection accuracy are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to a visual detection method for circuit board defects, an electronic device and a storage medium. Background Art

[0002] With the rapid development of science and technology, machine vision inspection has become a very common inspection tool in modern industrial production. Compared with the human eye, machine vision can better detect errors in the production process, better detect product quality problems, improve industrial production efficiency and production automation, and improve the accuracy of industrial production, improve, speed up work processes and save time.

[0003] In the existing technical solutions, visual inspection of circuit defects through AI visual inspection has become an indispensable inspection link in the production line of LED light strip circuit boards. However, the existing AI visual inspection method determines defects through AOI equipment and segments the circuit board image to obtain defect images, and then detects the defect images through pre-trained models. However, due to the omission and misjudgment of some complex defects by AOI equipment, it is impossible to obtain defect images of these complex defects. On the other hand, the segmented defect images cannot specifically show the location and size of the defects, resulting in the pre-trained model being unable to accurately identify these images when detecting them. Summary of the invention

[0004] The embodiments of the present application provide a visual inspection method for circuit board defects, an electronic device, and a storage medium, which improve the defect detection rate and inspection accuracy.

[0005] In the first aspect, the embodiment of the present application provides a visual detection method for circuit board defects, which is applied to an intelligent visual recognition platform, wherein the intelligent visual recognition platform includes a positioning module, a classification module, and a detection module, and the intelligent visual recognition platform is communicatively connected to an AOI device; the visual detection method includes:

[0006] Acquire point reporting data through the AOI device, wherein the point reporting data includes a defect image and defect information generated by the AOI device after taking a photo of the circuit board to be inspected, and the defect information includes a defect category existing on the circuit board to be inspected identified by the AOI device;

[0007] Determine the target area to be detected on the defect image by the positioning module;

[0008] Identifying, by the classification module, the category of the object to be detected contained in the target area, wherein the category of the object to be detected includes the type of device and / or the type of connection terminal on the circuit board;

[0009] The detection module locates a defective area with defects in the target area according to the category of the object to be detected and the defect category of the defect information, and outputs a detection result according to the position and size of the defective area.

[0010] In some embodiments, identifying the category of the object to be detected contained in the target area by the classification module includes:

[0011] According to the device database and the target area, a device standard map is obtained;

[0012] Identifying the category of the object to be detected contained in the target area according to the device standard map by the classification module;

[0013] The device database is used to store standard drawings of multiple devices.

[0014] In some embodiments, the locating a defective area having defects in the target area according to the category of the object to be detected and the defect category of the defect information by the detection module includes:

[0015] Extracting a feature vector from the target area according to the category of the object to be detected and the defect category of the defect information by the detection module to obtain a defect feature vector;

[0016] The defect region is located in the target region according to the defect feature vector.

[0017] In some embodiments, outputting the detection result according to the position and size of the defect area includes:

[0018] Acquire a defect template according to the defect database and the position and size of the defect area;

[0019] According to the defect template and the defect area, the detection result is obtained and outputted;

[0020] The defect database is used to store sample templates of various defects.

[0021] In some embodiments, the defect category of the defect information includes character printing errors, the intelligent visual recognition platform further includes an optical character recognition model, and after the classification module identifies the category of the object to be detected contained in the target area according to the device standard diagram, the visual inspection method further includes:

[0022] Recognize the character information on the object to be detected by using the optical character recognition model;

[0023] Character defect information is obtained according to the device standard diagram and the character information, and the character defect information is integrated into the detection result output.

[0024] In some embodiments, after outputting the detection result according to the position and size of the defective area, the method further comprises:

[0025] Displaying the test results and obtaining manual re-judgment results with the test results as a reference;

[0026] The positioning module, the classification module and the detection module are updated according to the detection result and the manual re-judgment result.

[0027] In some embodiments, the defect image and the defect information are generated based on an image of the circuit board to be inspected taken by the AOI device and a prefabricated standard image.

[0028] In some embodiments, the intelligent visual recognition platform further includes a monitoring module; the step of obtaining the reporting point data through the AOI device includes:

[0029] Monitoring the target folder for receiving the reporting point data through the monitoring module;

[0030] When the target folder receives the reporting point data, the reporting point data is read from the target folder, and the reporting point data is parsed to obtain the defect image and the defect information.

[0031] In a second aspect, an embodiment of the present application provides an intelligent visual recognition platform, including a memory and a processor, wherein the memory stores a computer program, and the processor implements the visual detection method described in the first aspect when executing the computer program.

[0032] In a third aspect, an embodiment of the present application provides a computer-readable storage medium, in which a computer program is stored. When the computer program is run on a computer, the computer executes the method described in the first aspect.

[0033] A visual inspection method, electronic device and storage medium for circuit board defects provided by the embodiments of the present application have at least the following beneficial effects: in the embodiments of the present application, reporting data including defect images and defect information are obtained by an AOI device, an area to be inspected is determined according to the defect image by a positioning module, types of various objects to be inspected in the area to be inspected are identified by a classification module, and the defect area is determined according to the various objects to be inspected and the defect information in the area to be inspected by a detection module, thereby locating the position and size of the defect, and identifying and outputting the detection information; therefore, the embodiments of the present application determine the area to be inspected by a positioning module, narrow the detection range and concentrate it in the area to be inspected, identify the types of various objects to be inspected in the area to be inspected by a classification module, and identify whether the objects to be inspected in the area to be inspected have defects according to different types of objects to be inspected and defect types in defect information by a detection model, further narrow the detection range and locate the defect area, thereby determining the position and size of the defect, greatly improving the defect detection rate and detection accuracy.

[0034] Other features and advantages of the present application will be described in the following description, and partly become apparent from the description, or understood by practicing the present application. The purpose and other advantages of the present application can be realized and obtained by the structures specifically pointed out in the description, claims and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 It is a flow chart of a visual inspection method for circuit board defects provided in an embodiment of the present application;

[0036] Figure 2 It is a flowchart of a method for identifying an object to be detected provided in an embodiment of the present application;

[0037] Figure 3 It is a flow chart of a method for locating a defective area provided in an embodiment of the present application;

[0038] Figure 4 It is a flowchart of a method for determining a test result provided in an embodiment of the present application;

[0039] Figure 5 It is a flowchart of a method for identifying character printing defects provided in an embodiment of the present application;

[0040] Figure 6 It is a flowchart of a re-judgment update method provided in an embodiment of the present application;

[0041] Figure 7 It is a flowchart of a data monitoring method provided in an embodiment of the present application;

[0042] Figure 8It is a structural diagram of an intelligent visual recognition platform provided in an embodiment of the present application;

[0043] Fig. 9 It is a schematic diagram of the overall process of a visual detection method for circuit board defects provided in an embodiment of the present application. DETAILED DESCRIPTION

[0044] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application is further described in detail below in conjunction with the accompanying drawings and examples. It should be understood that the specific embodiments described herein are only used to explain the present application and are not intended to limit the present application. In addition, the characteristics, operations or features described in the specification can be combined in any appropriate manner to form various implementation methods. At the same time, the steps or actions in the method description can also be replaced or adjusted in order in a manner that is obvious to those skilled in the art. Therefore, the various sequences in the specification and the accompanying drawings are only for the purpose of clearly describing a certain embodiment and are not meant to be a necessary sequence, unless otherwise specified that a certain sequence must be followed.

[0045] In the description of this application, "several" means one or more, "more" means more than two, "greater than", "less than", "exceed", etc. are understood to exclude the number itself, and "above", "below", "within", etc. are understood to include the number itself. If there is a description of "first" or "second", it is only used for the purpose of distinguishing technical features, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features or implicitly indicating the order of the indicated technical features.

[0046] The serial numbers of the components in this document, such as "first", "second", etc., are only used to distinguish the objects described and do not have any order or technical meaning. The "connection" and "coupling" mentioned in this application, unless otherwise specified, include direct and indirect connections (couplings).

[0047] With the rapid development of science and technology, visual inspection has become a very common inspection tool in modern industrial production. The characteristics of machine vision inspection are to improve the flexibility and automation of production. In some dangerous working environments that are not suitable for manual operation or where artificial vision cannot meet the requirements, machine vision is often used to replace artificial vision.

[0048] Compared with manual vision, visual inspection can better detect errors in the production process and better detect product quality problems. At the same time, in large-scale industrial production processes, using manual vision to check product quality is inefficient and not very accurate. Using machine vision inspection methods can greatly improve production efficiency and the degree of production automation, improve the accuracy of industrial production, speed up work processes and save time.

[0049] As the application of visual inspection becomes more and more extensive, the application scenarios are also becoming more and more diverse. In the existing technical solutions, the LED light strip circuit board is illuminated by light sources of different brightness, the image of the LED light strip circuit board is taken by an optical lens, and then the image of the LED light strip circuit board is analyzed and processed by a pre-trained model to determine the defects on the LED light strip circuit board. However, due to the omission and misjudgment of complex defects on different devices by AOI equipment, it is impossible to obtain defect images of these complex defects. On the other hand, the defect images obtained by the AOI equipment cannot specifically show the location and size of the defects, and it is impossible to accurately and appropriately locate and select the required detection range, resulting in the pre-trained model being unable to detect these images and make accurate identification.

[0050] Based on this, the embodiments of the present application provide a visual detection method, electronic device and storage medium for circuit board defects, which improve the defect detection rate and detection accuracy.

[0051] The following is a description of a visual inspection method for circuit board defects, an electronic device, and a storage medium provided in an embodiment of the present application in conjunction with the accompanying drawings:

[0052] Reference Figure 1 As shown, Figure 1 : is a flow chart of a visual detection method for circuit board defects provided by an embodiment of the present application. The method may include but is not limited to the following steps:

[0053] Step S110, obtaining point reporting data through the AOI device, the point reporting data including defect images and defect information generated after the AOI device takes a picture of the circuit board to be inspected, and the defect information including the defect category existing on the circuit board to be inspected identified by the AOI device;

[0054] It should be noted that the AOI equipment performs high-resolution scanning or photography of the LED light strip circuit board to obtain images of the entire board surface. These images can clearly show the various components and solder joints on the circuit board. The AOI equipment pre-processes the captured images through built-in software, such as grayscale conversion, filtering, edge enhancement, etc., to improve the identifiability of defects. Once the defects are identified and classified, the system will generate a point reporting data containing defect images and detailed defect information. The point reporting data can include the number of defect images, the defect type, defect number and defect location identified by the defect image and the AOI equipment.

[0055] In one embodiment of the present application, the AOI device performs high-resolution scanning or photography of the LED light strip circuit board, and compares the captured image with a pre-made LED light strip circuit board template. If the comparison result is defective, the captured image is pre-processed by the built-in software of the AOI device to obtain a defective image and generate point reporting data. The intelligent visual recognition platform is connected to the AOI device for communication and sends the point reporting data to the intelligent visual recognition platform for AI visual inspection.

[0056] It is understandable that the AOI equipment can automatically identify pad defects, pin misalignment, and welding short circuit effects, conduct preliminary inspections on LED light strip circuit boards, and obtain the reporting data of the LED light strip circuit boards that are identified and generated by the AOI equipment, which is convenient for subsequent AI visual inspections through the intelligent visual recognition platform.

[0057] Step S120, determining a target area to be detected on the defect image by a positioning module;

[0058] It should be noted that the positioning module includes a pre-trained positioning model, which is used to identify the target area to be detected in the defect image. The defect image can have one or more defects. The defects on the defect image are divided into regions through the pre-trained positioning model, and the target area to be detected with defects including all defects on the defect image and an appropriate area size is segmented, so that subsequent defect recognition only needs to be performed in the target area to be detected.

[0059] In one example of the present application, a defect image is obtained based on the reported point data, the defect location is identified based on the defect image through a pre-trained positioning model, and a target area to be detected that includes all defects and has an appropriate size is output.

[0060] It is understandable that determining the target area through the positioning model can help narrow the detection range. After determining the target area, only these areas can be detected in detail, reducing unnecessary calculations and processing, and improving the efficiency and speed of detection. On the other hand, determining the target area can make the detection algorithm focus more on important areas, thereby improving the accuracy and precision of detection and reducing the possibility of false alarms and missed detections.

[0061] Step S130, identifying the category of the object to be detected contained in the target area by a classification module, the category of the object to be detected includes the type of device and / or the type of connection terminal on the circuit board;

[0062] It should be noted that the classification module includes a pre-trained classification model, which is used to identify the categories of all objects to be detected in the target area, such as various types of electronic components such as resistors, capacitors, transistors, diodes, integrated circuits, and conductive parts such as solder joints, pins, and sockets that connect electronic components.

[0063] It is understandable that identifying different device types and connection terminal types helps to further identify potential defects. Corresponding device types and connection terminal types will have corresponding special defects, which will only appear on the corresponding device types and / or connection terminal types. By identifying different device types and connection terminal types, these special defects can be more effectively discovered. On the other hand, identifying different device types and connection terminal types helps to further identify potential defects, such as finding mis-installed parts or welding anomalies.

[0064] Step S140 , the detection module locates a defective area in the target area according to the category of the object to be detected and the defect category of the defect information, and outputs a detection result according to the location and size of the defective area.

[0065] It should be noted that the detection module includes a pre-trained detection model, which is used to detect and locate defective areas in the target area, identify defects in the area to be detected according to the category of the object to be detected and the defect category of the defect information, and locate the defective area, thereby obtaining the position and size of the defective area.

[0066] In one embodiment of the present application, the detection module detects whether the category of each object to be detected corresponds to the defect category according to the category of the object to be detected and the defect category, locates the position and size of the defect area, reduces the defect area to the location of the defect on the detection object, determines the specific information of a certain defect of a certain detection object, and outputs it as the detection result.

[0067] It can be understood that the detection module can accurately locate the position of defects in the target area, help production personnel quickly find the problem, and improve the accuracy and efficiency of problem location. The automated detection module can efficiently scan the target area and identify which inspection object the defective area is located on, effectively improving the defect detection efficiency on the production line, reducing costs and improving product quality.

[0068] Reference Figure 2 As shown, Figure 2 : is a flow chart of a method for identifying an object to be detected provided by an embodiment of the present application. The method may include but is not limited to the following steps:

[0069] Step S210, obtaining a device standard diagram according to the device database and the target area;

[0070] The device database is used to store standard drawings of multiple devices.

[0071] It should be noted that the device database is used to store standard drawings of multiple devices, and the standard drawings of each device should include information such as its shape, characteristics, and connection methods for subsequent classification and identification. The positioning module retrieves the standard drawings of similar devices based on the degree of similarity between each object to be detected in the target area to be detected and the device standard drawings in the device database.

[0072] Step S220: identifying the category of the object to be detected contained in the target area according to the device standard map by a classification module.

[0073] It should be noted that the categories of the objects to be detected contained in the target area are identified one by one according to the retrieved device standard diagram. The classification module needs to be able to effectively compare and match the target area data and the device standard diagram. Identifying different device types and connection terminal types helps to further identify potential defects, such as mis-installed components or welding anomalies, to achieve accurate identification and classification. The automated matching algorithm can quickly process a large number of areas to be detected and generate corresponding test results, thereby improving the efficiency of detection.

[0074] Reference Figure 3 As shown, Figure 3 : is a flow chart of a method for locating a defective area provided in an embodiment of the present application. The method may include but is not limited to the following steps:

[0075] Step S310, extracting a feature vector from the target area according to the category of the object to be detected and the defect category of the defect information by the detection module to obtain a defect feature vector;

[0076] It should be noted that, according to the category of the object to be detected and the defect category of the defect information, the detection module will design corresponding feature extraction methods for different defect types. In an embodiment of the present application, the convolutional neural network backbone in the detection module is used to extract features of the target area to be detected. These features can be information such as edges, textures, colors, shapes, etc. in the image data, or certain attributes or features in other data types. The features are statistically analyzed, transformed, reduced in dimension, or processed by other methods to ensure that the features can accurately describe the characteristics of the target area, and the described features are converted into the form of feature vectors. Normally, a feature vector is a vector composed of features, and the position of each feature in the vector corresponds to its position in the original data, for subsequent defect identification, positioning, and analysis.

[0077] Step S320: locating the defect area in the target area according to the defect feature vector.

[0078] It should be noted that the feature vector of the object to be detected is matched with the known defect feature vector, and the similarity measure between the feature vectors is analyzed to identify possible defect feature vectors. Generally, the area corresponding to the feature vector with high defect similarity is more likely to be a defect area. Based on the results of the similarity analysis, the possible defect feature vector is determined, and the corresponding area is marked as a potential defect area. The potential defect area is further verified and analyzed to confirm whether there is a defect.

[0079] Reference Figure 4 As shown, Figure 4 : is a flow chart of a method for determining a test result provided in an embodiment of the present application, and the method may include but is not limited to the following steps:

[0080] Step S410, obtaining a defect template according to the defect database and the position and size of the defect area;

[0081] Among them, the defect database is used to store sample templates of various defects.

[0082] It should be noted that, in one embodiment of the present application, the detection module matches a similar defect template from a defect database according to the position and size of the defect area. In another embodiment of the present application, the defect template in the defect database is retrieved and the feature vector of the defect template and the area to be detected is compared to find the best matching defect template.

[0083] Step S420, obtaining and outputting the detection result according to the defect template and the defect area.

[0084] It should be noted that the type, location, size and other information of the defects in the area to be detected are determined by matching the defect template, and the matched defect template is superimposed or compared with the area to be detected to obtain the detection result.

[0085] It can be understood that by matching the defect database and the area to be inspected, the type, location and size of the defect can be accurately determined, thereby improving the accuracy of the inspection result. The defect database can be continuously updated and expanded to meet the inspection needs in different scenarios, thereby improving the applicability and scalability of the method.

[0086] Reference Figure 5 As shown, Figure 5 : is a flow chart of a method for identifying character printing defects provided in an embodiment of the present application. The method may include but is not limited to the following steps:

[0087] Step S510, recognizing character information on the object to be detected by using an optical character recognition model;

[0088] It should be noted that the character information on the object to be inspected is recognized through the trained optical character recognition model, and the characters in the image are converted into text form, so as to identify character printing defects.

[0089] Step S520, obtaining character defect information according to the device standard diagram and the character information, and integrating the character defect information into the detection result output.

[0090] It should be noted that, based on the device standard diagram and the recognized character information, character information related to the device standard is extracted, and defect recognition is performed on the extracted character information to detect possible errors, missing or damaged defects therein, which can be achieved by comparing the extracted character information with the standard character template or rule. The recognized character defect information is integrated with other test results, and the final test result is output.

[0091] It is understandable that compared with manual processing of character information, the optical character recognition model can reduce human interference and errors, improve recognition accuracy, and reduce the misjudgment rate of defect recognition. On the other hand, by matching the device standard image and character information, it is possible to achieve standardized processing of character defects, ensure consistent judgment of defects, and improve the reliability of detection results.

[0092] Reference Figure 6 As shown, Figure 6 : is a flowchart of a re-judgment update method provided in an embodiment of the present application, and the method may include but is not limited to the following steps:

[0093] Step S610, displaying the test results, and obtaining manual re-judgment results with the test results as a reference;

[0094] Step S620, updating the positioning module, classification module and detection module according to the detection results and manual re-judgment results.

[0095] It should be noted that the test results are displayed to the equipment of the re-inspection personnel, and the re-inspection workers re-judge the LED light strip circuit board based on the test results to determine the manual re-judgment results. When the manual re-judgment results are consistent with the test results, the device database and defect database are updated according to the test results, so as to identify more LED light strip circuit board defects in subsequent use; when the manual re-judgment results are inconsistent with the test results, the pre-trained positioning model in the positioning module, the pre-trained classification model in the classification module, and the pre-trained detection model in the detection module are updated according to the manual re-judgment results and the test results, and finally the device database and defect database are updated.

[0096] In some embodiments, the defect image and defect information are generated based on an image of the circuit board to be inspected taken by the AOI device and a prefabricated standard image.

[0097] It should be noted that for the circuit board to be inspected, a corresponding pre-made standard diagram is prepared, and images of the circuit board to be inspected with different brightness are obtained through the AOI equipment. The captured image is compared with the pre-made standard diagram. The AOI equipment can generate a defect image and corresponding defect information. The defect information includes information such as the type, location, and size of the defect.

[0098] Reference Figure 7 As shown, Figure 7 : is a flow chart of a data monitoring method provided by an embodiment of the present application, and the method may include but is not limited to the following steps:

[0099] Step S710, monitoring the target folder for receiving the reporting point data through the monitoring module;

[0100] Step S720, when the target folder receives the point reporting data, the point reporting data is read from the target folder, and the point reporting data is parsed to obtain the defect image and defect information.

[0101] It should be noted that after the AOI device generates a defect image and the corresponding defect information, the AOI device integrates the defect image and the corresponding defect information into point reporting data, and sends it to the target folder for receiving the point reporting data on the intelligent visual recognition platform. The monitoring module monitors the information that the target folder receives the point reporting data, reads the point reporting data from the target folder, and parses the point reporting data to obtain the defect image and defect information.

[0102] It can be understood that the monitoring module can monitor the target folder in real time, promptly sense the arrival of the reporting data, and thus promptly trigger the data processing process to ensure the timely processing of the reporting data. By continuously monitoring the target folder, missed detections caused by failure to timely process the reporting data can be avoided, ensuring that each reporting data is processed and recorded in a timely manner.

[0103] Reference Figure 8 As shown, Figure 8 It is a structural diagram of an intelligent visual recognition platform provided in an embodiment of the present application.

[0104] The intelligent visual recognition platform includes:

[0105] The processor 810 may be implemented by a general-purpose central processing unit (CPU), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant instructions to implement the technical solutions provided in the embodiments of the present application;

[0106] The memory 820 can be implemented in the form of a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 820 can store an operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented by software or firmware, the relevant program codes are stored in the memory 820, and the processor 810 calls and executes the technical solutions of the embodiments of this application;

[0107] Input / output interface 830, used to implement information input and output;

[0108] The communication interface 840 is used to realize the communication interaction between the device and other devices. The communication can be realized through a wired manner (such as USB, network cable, etc.) or a wireless manner (such as mobile network, WIFI, Bluetooth, etc.);

[0109] bus 850 , which transmits information between the various components of the device (e.g., processor 810 , memory 820 , input / output interface 830 , and communication interface 840 );

[0110] The processor 810 , the memory 820 , the input / output interface 830 and the communication interface 840 are connected to each other in communication within the device via a bus 850 .

[0111] The embodiment of the present application further provides a computer-readable storage medium, in which a computer program is stored. When the computer program runs on a computer, the computer executes the method described above.

[0112] Reference Fig. 9 As shown, Fig. 9 This is a schematic diagram of the overall process of a visual inspection method for circuit board defects provided by an embodiment of the present application. The method may include but is not limited to the following steps:

[0113] Step S901, obtaining the reporting point data through the AOI device;

[0114] Step S902, monitoring the target folder for receiving the reporting point data through the monitoring module;

[0115] Step S903, when the target folder receives the point reporting data, the point reporting data is read from the target folder, and the point reporting data is parsed to obtain the defect image and defect information;

[0116] Step S904, determining the target area to be detected on the defect image by using a positioning module;

[0117] Step S905, obtaining a device standard diagram according to the device database and the target area;

[0118] Step S906, identifying the category of the object to be detected contained in the target area according to the device standard diagram by a classification module;

[0119] Step S907, extracting a feature vector from the target area according to the category of the object to be detected and the defect category of the defect information by the detection module to obtain a defect feature vector;

[0120] Step S908, locating the defect area in the target area according to the defect feature vector;

[0121] Step S909, obtaining a defect template according to the defect database and the position and size of the defect area;

[0122] Step S910, obtaining and outputting the detection result according to the defect template and the defect area.

[0123] In one embodiment of the present application, a corresponding prefabricated standard image is prepared for a circuit board to be inspected, and images of the circuit board to be inspected with different brightness are obtained through an AOI device, and the captured image is compared with the prefabricated standard image. If the comparison result is a defect, the AOI device can generate a defect image and corresponding defect information, and the defect information includes information such as the type, location, and size of the defect. The intelligent visual recognition platform is connected to the AOI device for communication, and the reporting data is sent to the intelligent visual recognition platform for AI visual inspection. The monitoring module monitors the information that the target folder receives the reporting data, reads the reporting data from the target folder, parses the reporting data to obtain a defect image and defect information, obtains a defect image based on the reporting data, identifies the defect location based on the defect image through a pre-trained positioning model, and outputs a target area to be inspected that includes all defects and has an appropriate size. The positioning module retrieves relatively similar devices based on the degree of similarity between each object to be inspected in the target area to be inspected and the device standard image in the device database. Standard map, according to the retrieved device standard map, the categories of the objects to be detected contained in the target area are identified one by one, the convolutional neural network backbone in the detection module is used to extract features of the target area to be detected, and the features are statistically analyzed, transformed, reduced in dimension or other processing methods are used to ensure that the features can accurately describe the characteristics of the target area, and the described features are converted into the form of feature vectors, and the feature vectors of the objects to be detected are matched with the known defect feature vectors. According to the results of the similarity analysis, the possible defect feature vectors are determined, and the corresponding areas are marked as potential defect areas. The potential defect areas are further verified and analyzed to confirm whether there are really defects. This is achieved by retrieving the defect template in the defect database and comparing the defect template with the feature vector of the area to be detected to find the best matching defect template. The defect type, location, size and other information in the area to be detected are determined through the matched defect template, and the matched defect template is superimposed or compared with the area to be detected to obtain the detection result.

[0124] In an embodiment of the present application, the reporting data including the defect image and the defect information is obtained through the AOI device, the area to be detected is determined according to the defect image through the positioning module, the types of various objects to be detected in the area to be detected are identified through the classification module, and the defect area is determined according to the various objects to be detected and the defect information in the area to be detected by the detection module, thereby locating the position and size of the defect, identifying and outputting the detection information; therefore, the embodiment of the present application determines the area to be detected through the positioning module, narrows the detection range and concentrates it in the area to be detected, identifies the types of various objects to be detected in the area to be detected through the classification module, and identifies whether the objects to be detected in the area to be detected have defects according to the different types of objects to be detected and the defect types in the defect information through the detection model, further narrows the detection range and locates the defect area, thereby determining the position and size of the defect, greatly improving the defect detection rate and detection accuracy.

[0125] The terms "first", "second", "third", "fourth", etc. (if any) in the specification of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0126] It should be understood that in the present application, "at least one (item)" means one or more, and "plurality" means two or more. "And / or" is used to describe the association relationship of associated objects, indicating that three relationships may exist. For example, "A and / or B" can mean: only A exists, only B exists, and A and B exist at the same time, where A and B can be singular or plural. The character " / " generally indicates that the objects associated before and after are in an "or" relationship. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, c can be single or multiple.

[0127] It should also be understood that the various implementations provided in the embodiments of the present application can be combined arbitrarily to achieve different technical effects.

[0128] The above is a specific description of the preferred implementation of the present application, but the present application is not limited to the above-mentioned implementation mode. Technical personnel familiar with the field can also make various equivalent deformations or substitutions without violating the spirit of the present application. These equivalent deformations or substitutions are all included in the scope defined by the claims of the present application.

Claims

1. A visual inspection method for circuit board defects, characterized in that: Applied to an intelligent visual recognition platform, the intelligent visual recognition platform includes a positioning module, a classification module and a detection module, and the intelligent visual recognition platform is communicatively connected with an AOI device; the visual detection method includes: Acquire point reporting data through the AOI device, wherein the point reporting data includes a defect image and defect information generated by the AOI device after taking a photo of the circuit board to be inspected, and the defect information includes a defect category existing on the circuit board to be inspected identified by the AOI device; Determine the target area to be detected on the defect image by the positioning module; Identifying, by the classification module, the category of the object to be detected contained in the target area, wherein the category of the object to be detected includes the type of device and / or the type of connection terminal on the circuit board; The detection module locates a defective area with defects in the target area according to the category of the object to be detected and the defect category of the defect information, and outputs a detection result according to the position and size of the defective area.

2. The visual inspection method according to claim 1, characterized in that: The identifying, by the classification module, the category of the object to be detected contained in the target area comprises: According to the device database and the target area, a device standard map is obtained; Identifying the category of the object to be detected contained in the target area according to the device standard map by the classification module; The device database is used to store standard drawings of multiple devices.

3. The visual inspection method according to claim 1, characterized in that: The method of locating a defective area in the target area according to the category of the object to be detected and the defect category of the defect information by the detection module includes: Extracting a feature vector from the target area according to the category of the object to be detected and the defect category of the defect information by the detection module to obtain a defect feature vector; The defect region is located in the target region according to the defect feature vector.

4. The visual inspection method according to claim 1, characterized in that: Outputting the detection result according to the position and size of the defective area includes: Acquire a defect template according to the defect database and the position and size of the defect area; According to the defect template and the defect area, the detection result is obtained and outputted; The defect database is used to store sample templates of various defects.

5. The visual inspection method according to claim 2, characterized in that: The defect category of the defect information includes character printing errors, the intelligent visual recognition platform also includes an optical character recognition model, and after the classification module identifies the category of the object to be detected contained in the target area according to the device standard diagram, the visual inspection method further includes: Recognize the character information on the object to be detected by using the optical character recognition model; Character defect information is obtained according to the device standard diagram and the character information, and the character defect information is integrated into the detection result output.

6. The visual inspection method according to claim 1, characterized in that: After outputting the detection result according to the position and size of the defective area, the method further comprises: Displaying the test results and obtaining manual re-judgment results with the test results as a reference; The positioning module, the classification module and the detection module are updated according to the detection result and the manual re-judgment result.

7. The visual inspection method according to claim 1, characterized in that: The defect image and the defect information are generated according to an image of the circuit board to be inspected taken by the AOI device and a prefabricated standard image.

8. The visual inspection method according to claim 1, characterized in that: The intelligent visual recognition platform also includes a monitoring module; the step of obtaining the reporting point data through the AOI device includes: Monitoring the target folder for receiving the reporting point data through the monitoring module; When the target folder receives the reporting point data, the reporting point data is read from the target folder, and the reporting point data is parsed to obtain the defect image and the defect information.

9. An intelligent visual recognition platform, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the visual inspection method according to any one of claims 1 to 8 when executing the computer program.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed on a computer, the computer is enabled to execute the method according to any one of claims 1 to 8.

Citation Information

Patent Citations

  • Appearance detection method and device of camera module, electronic equipment and storage medium

    CN110415214A

  • Two-dimensional PCB appearance defect real-time automatic detection technology based on deep learning

    CN111754456A

  • Image detection method and device, equipment and storage medium

    CN114820457A

  • Defect detection method, defect detection device and computer readable storage medium

    CN115249237A

  • Method and apparatus for detecting defective markings on a semiconductor product

    US20020092910A1

Cited By

  • Industrial product quality detection method and system based on machine vision

    CN119600032A

  • Battery detection method, battery detection equipment, device, storage medium and product

    CN121504942A