Automatic optical detection device for PCB (Printed Circuit Board) defects based on machine vision

Abstract

The invention discloses an automatic optical detection device for PCB (Printed Circuit Board) defects based on machine vision, relates to the technical field of PCB defect detection, solves the technical problem in the prior art that targeted defect identification detection cannot be carried out on a background area, a mutual matching area and a component point location, and particularly relates to an illumination influence evaluation unit. Performing influence evaluation on the lighting module; after no influence is evaluated, the lighting module is put into use, and image acquisition is carried out on the PCB; the image background separation unit is used for processing the acquired image, namely separating the image background, and obtaining a background region, a mutual matching region and a component point location through separation; the processing trace analysis unit is used for carrying out processing trace analysis on the mutual matching areas and the component point positions; and the defect identification and detection unit is used for carrying out defect identification and detection on the background area, the mutual matching area and the component point positions.

Description

Technical Field

[0001] The present invention relates to the technical field of PCB board defect detection, and specifically to an automatic optical detection device for PCB board defects based on machine vision. Background Technique

[0002] Automatic optical detection of PCB board defects is a method that uses optical imaging technology and image processing algorithms to automatically detect various defects on the PCB board; the image of the PCB board is collected by a high-resolution camera, and then the image is transmitted to a computer for processing; the computer uses image processing algorithms to analyze the image and extract the characteristic information on the PCB board, such as circuits, solder joints, components, etc.; these characteristic information are compared with the pre-set standard templates or rules to determine whether there are defects and the type and location of the defects.

[0003] However, in the prior art, it is impossible to separate the PCB background and components, resulting in inaccurate defect warning during defect recognition, there is a risk of unnecessary defect warning, which affects the PCB processing progress. In addition, it is impossible to perform targeted defect recognition and detection on the background area, the mating area, and the component points, reducing the comprehensiveness of PCB board defect detection.

[0004] In view of the above technical defects, a solution is now proposed. Summary of the Invention

[0005] The purpose of the present invention is to solve the above-mentioned problems and propose an automatic optical detection device for PCB board defects based on machine vision.

[0006] The purpose of the present invention can be achieved by the following technical solutions:

[0007] An automatic optical detection device for PCB board defects based on machine vision includes an optical detection center, an illumination module, an image acquisition module, and a motion control module; the optical detection center is communicatively connected to:

[0008] An illumination impact assessment unit that assesses the impact on the illumination module; after the assessment shows no impact, the illumination module is put into use and the image of the PCB board is collected;

[0009] An image background separation unit that processes the collected image, that is, separates the image background, and obtains the background area, the mating area, and the component points through separation;

[0010] A processing trace analysis unit that analyzes the processing traces of the mating area and the component points;

[0011] A defect recognition and detection unit that performs defect recognition and detection on the background area, the mating area, and the component points.

[0012] As a preferred embodiment of the present invention, the process of the lighting impact assessment unit is as follows:

[0013] When taking the light source position saved in any setting of the lighting module as the real-time light source position, perform lighting analysis on all components, collect the mean deviation of the marked luminous flux in the near-light area of components at different positions, and at the same time collect the average value of the continuous decrease span of the luminous flux in the far-light area of components at different positions, and perform threshold comparison:

[0014] If the mean deviation of the luminous flux exceeds the threshold of the mean deviation of the luminous flux, or the average value of the continuous decrease span of the luminous flux exceeds the threshold of the continuous decrease span mean, then generate a high lighting impact signal; if the mean deviation of the luminous flux does not exceed the threshold of the mean deviation of the luminous flux, and the average value of the continuous decrease span of the luminous flux does not exceed the threshold of the continuous decrease span mean, then generate a low lighting impact signal and send it to the optical detection center and the lighting module.

[0015] As a preferred embodiment of the present invention, the process of the image background separation unit is as follows:

[0016] Perform frame-by-frame picture analysis on the collected images, identify all the collected pictures, obtain the on-picture positions of the components in the PCB through the image ratio, and mark the positions of the components in the collected pictures; classify the component types into independently operating components and cooperatively operating components according to the working mode type; mark the mutual cooperation areas of the cooperatively operating components; collect the image parameters of the collected pictures; mark the points with deviations from the image parameters of the surrounding positions, and compare them with the on-picture positions of the components. If the marked points overlap with the on-picture positions of the components, mark the surrounding positions of the corresponding marked points as the background area; and correspond the marked points with the on-picture positions of the components one by one, and at the same time collect the mutual cooperation areas.

[0017] As a preferred embodiment of the present invention, during the comparison process, if the marked points cannot correspond to the on-picture positions of the corresponding components, set the corresponding marked points as defect risk points, and compare the image parameters of the defect risk points;

[0018] If the image parameters of the defect risk points exceed the set image parameter range of the background area, it is inferred that there are defects at the defect risk points; mark the currently collected picture as a defect-generated picture, and perform defect change detection based on adjacent collected pictures. When the defect appears and the defect area range increases, generate a background area defect signal; when the defect does not change and the components adjacent to the defect risk points operate normally, generate a background area wear signal.

[0019] As a preferred embodiment of the present invention, if the image parameters of the defect risk point do not exceed the set image parameter range of the background area, it is inferred that there is a defect risk at the defect risk point, and the collected pictures at adjacent historical moments are analyzed starting from the current moment. If the image parameters of the defect risk point in the collected pictures corresponding to adjacent moments float numerically in a trend direction that exceeds the image parameter range, a defect warning signal is generated;

[0020] If the image parameters of the defect risk point in the collected pictures corresponding to adjacent moments do not float numerically in a trend direction that exceeds the image parameter range, a normal wear signal of the sheet is generated and the normal wear signal of the sheet is sent to the optical detection center.

[0021] As a preferred embodiment of the present invention, the process of the processing trace analysis unit is as follows:

[0022] The area where the image parameters of the component point are deviated from the image parameters of the background area is marked as the point welding area; the deviation distance between the boundary of the point welding area and the covered area of the component spot welding method is collected; the boundary of the line track and the actual line welding track in the mutually matching area are collected, and the boundary offset of the track is obtained according to the comparison of the track boundaries.

[0023] As a preferred embodiment of the present invention, the deviation distance between the boundary of the point welding area and the covered area of the component spot welding method and the boundary offset of the track are respectively compared with the deviation distance value and the offset threshold:

[0024] If the deviation distance between the boundary of the point welding area and the covered area of the component spot welding method exceeds the deviation distance value, or the boundary offset of the track exceeds the offset threshold, a processing influence signal is generated; if the deviation distance between the boundary of the point welding area and the covered area of the component spot welding method does not exceed the deviation distance value, and the boundary offset of the track does not exceed the offset threshold, a normal processing signal is generated and the normal processing signal is sent to the optical detection center.

[0025] As a preferred embodiment of the present invention, the defect identification and detection unit performs defect identification and detection on the mutually matching area and the component points; the position of the processing influence is collected and the position of the processing influence is marked on the collected picture. If the operation of the component corresponding to the starting moment of the appearance of the processing influence position is not affected, it is marked as deviation processing; and the change detection of the processing influence position of the deviation processing is performed; if the operation of the component corresponding to the starting moment of the appearance of the processing influence position is affected, it is marked as abnormal processing, and the processing welding method is adjusted and the processing influence position is repaired at the same time.

[0026] As a preferred embodiment of the present invention, the change detection is as follows: if the parameters of the processing influence position in adjacent acquired images show fluctuations, or the operating parameters of the components fluctuate in an abnormal trend, then the processing influence position is marked as the position affected by defects and sent to the optical detection center, and the optical detection center repairs the position affected by defects; if the parameters of the processing influence position in adjacent acquired images do not show fluctuations, and the operating parameters of the components do not fluctuate in an abnormal trend, then the processing influence position is marked as the position not affected by defects and sent to the optical detection center.

[0027] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0028] 1. In the present invention, the lighting module is evaluated for influence to ensure that the image acquisition environment during optical detection can guarantee the lighting conditions, avoid that the acquired images during optical detection cannot meet the actual image recognition requirements, and cannot effectively perform image recognition, thereby reducing the accuracy of optical detection.

[0029] 2. In the present invention, the PCB background and components in the acquired images are separated to facilitate the effective positioning of defect positions in the analysis of the acquired images, avoid the lack of pertinence in image analysis, and be unable to infer whether the abnormal image position is the PCB position or the component position, so that some image defects have no actual influence but are determined to be defective.

[0030] 3. In the present invention, the background area is detected for defects to ensure the usability of the PCB board area, avoid affecting the set operating efficiency of the components, and at the same time, the processing process can be effectively detected based on the defect detection of the background area, avoid that the processing process affects the PCB board surface and there are continuous influences during the PCB board processing, so as to reduce the use efficiency of the PCB and increase the risk of operating failures of the components in the PCB.

[0031] 4. In the present invention, the processing traces of the mutually cooperating areas and component points are analyzed to infer whether there are defect influences on the processing of the components in the PCB, so that the defects caused by the component processing on the PCB board affect the feasibility of the circuit operation in the entire PCB board. By comparing the traces, the accuracy of optical detection can be improved, and further improve the accuracy of defect detection and defect traceability analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] For the convenience of those skilled in the art to understand, the present invention will be further described below with reference to the accompanying drawings.

[0033] Figure 1 is the principle block diagram of the present invention;

[0034] Figure 2 is the principle block diagram of Embodiment 1 in the present invention;

[0035] Figure 3 This is the principle block diagram of Embodiment 2 in the present invention. Specific Embodiments

[0036] In order to enable those skilled in the art to better understand the solution of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without making creative efforts belong to the scope of protection of the present invention.

[0037] Referring to "embodiments" herein means that the specific features, structures, or characteristics described in connection with the embodiments can be included in at least one embodiment of the present invention. The phrase appears in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment mutually exclusive with other embodiments. Those skilled in the art explicitly and implicitly understand that the embodiments described herein can be combined with other embodiments.

[0038] Please refer to Figure 1 As shown, an automatic optical inspection device for PCB board defects based on machine vision includes an optical inspection center, wherein the optical inspection center is communicatively connected to an illumination module, an image acquisition module, and a motion control module; the PCB is subjected to defect detection by means of machine vision technology, and the illumination module is used to play an illumination role during the optical inspection process to ensure the image acquisition environment; the illumination module is composed of devices such as fill lights in the prior art;

[0039] The image acquisition module is used to acquire images of the PCB board, specifically including devices such as cameras, lenses, and lens light sources; the motion control module is used to control the movement of the camera;

[0040] Embodiment 1

[0041] Please refer to Figure 2 As shown, an illumination impact evaluation unit is communicatively connected between the optical inspection center and the illumination module. This embodiment is used to evaluate the impact on the illumination module to ensure that the image acquisition environment can guarantee the illumination conditions during the optical inspection process, avoid the situation that the acquired images during optical inspection cannot meet the actual image recognition requirements, and cannot effectively perform image recognition, thereby reducing the accuracy of optical inspection;

[0042] The optical inspection center generates an illumination impact evaluation signal and sends the illumination impact evaluation signal to the illumination impact evaluation unit. After receiving the illumination impact evaluation signal, the illumination impact evaluation unit evaluates the impact on the illumination module;

[0043] The PCBs that have been adaptively processed are correspondingly divided into several sub-regions, and the light source positions are set according to the positions of the corresponding sub-regions. With the light source positions saved in the current lighting module, when the current light source positions supply light, the components installed in the PCB are positioned and marked. The component positions and the light source positions divide the surrounding areas of the components into a near-light area and a far-light area;

[0044] When any of the light source positions saved in the lighting module is used as the real-time light source position, lighting analysis is performed on all components. The mean deviation of the marked light flux in the near-light areas of components at different positions is collected, and at the same time, the average value of the continuous decrease span of the light flux in the far-light areas of components at different positions is collected. The mean deviation of the marked light flux in the near-light areas of components at different positions and the average value of the continuous decrease span of the light flux in the far-light areas of components at different positions are respectively compared with the light flux mean deviation threshold and the continuous decrease span mean threshold:

[0045] If the mean deviation of the marked light flux in the near-light areas of components at different positions exceeds the light flux mean deviation threshold, or the average value of the continuous decrease span of the light flux in the far-light areas of components at different positions exceeds the continuous decrease span mean threshold, it is inferred that the current lighting impact assessment is abnormal, a lighting high-impact signal is generated and sent to the optical detection center and the lighting module; the lighting module re-screens and selects the currently saved light source positions. If the optical detection center is still affected by the lighting high-impact signal, the lighting module re-adjusts the light source specifications or the height of the light source positions;

[0046] If the mean deviation of the marked light flux in the near-light areas of components at different positions does not exceed the light flux mean deviation threshold, and the average value of the continuous decrease span of the light flux in the far-light areas of components at different positions does not exceed the continuous decrease span mean threshold, it is inferred that the current lighting impact assessment is normal, a lighting low-impact signal is generated and sent to the optical detection center and the lighting module;

[0047] Embodiment 2

[0048] After determining that there is no impact on the lighting impact assessment in the previous embodiment, this embodiment performs defect detection on the PCB. Please refer to Figure 3 As shown, in addition to the image acquisition module and the motion control module, the optical detection center is communicatively connected with an image background separation unit, a processing trace analysis unit, and a defect recognition and detection unit;

[0049] According to the light source positions saved in the lighting module, the PCB is supplemented with light and an image of the PCB is acquired. After the image acquisition is completed, the image acquisition is processed;

[0050] An image background separation unit is used to process the acquired image, separating the PCB background and components in the acquired image, so that the defect location can be effectively located in the analysis of the acquired image, avoiding the lack of pertinence in image analysis and being unable to infer whether the abnormal image location is the PCB location or the component location, so that some image defects have no actual impact but are determined to be defective, such as the surface wear of the PCB board causing a change in surface brightness but no defect is generated;

[0051] Perform frame-by-frame picture analysis on the acquired image, identify all the acquired pictures, obtain the on-picture location of the components in the PCB through the image ratio, and mark the component locations in the acquired pictures; and classify the component types into independently operating components and cooperating operating components according to the working mode type; and mark the mutual cooperation areas of the cooperating operating components; the mutual cooperation area is defined as the area where the components are related and cooperatively connected. For example, if two components communicate through a circuit, the area where the circuit is located is the mutual cooperation area;

[0052] Collect image parameters of the acquired pictures, such as clarity, brightness, colorfulness, etc.; mark the points with deviations from the image parameters of the surrounding positions, and compare them with the on-picture locations of the components. If the marked points overlap with the on-picture locations of the components, mark the surrounding positions of the corresponding marked points as background areas; and establish a one-to-one correspondence between the marked points and the on-picture locations of the components, and at the same time collect the mutual cooperation areas; that is, the real-time acquired picture consists of the background area, the mutual cooperation area, and the component points;

[0053] After the image background is separated, the defect recognition and detection unit performs defect detection on the background area to ensure the usability of the PCB board area, avoid affecting the set operating efficiency of the components, and at the same time, based on the defect detection of the background area, the processing process can be effectively detected, avoiding the processing process affecting the PCB board surface and having a continuous impact during the PCB board processing, so that the use efficiency of the PCB is reduced and the risk of component operation failure in the PCB is increased;

[0054] During the comparison process, if the marked points cannot correspond to the on-picture locations of the corresponding components, set the corresponding marked points as defect risk points, and compare the image parameters of the defect risk points;

[0055] If the image parameters of the defect risk points exceed the set image parameter range of the background area, it is inferred that there are defects at the defect risk points; the currently collected picture is marked as a defect-generated picture, and defect change detection is performed based on adjacent collected pictures. When the defect changes and the range of the defect area increases, a background area defect signal is generated and sent to the optical detection center. After receiving the background area defect signal, the optical detection center repairs the PCB board surface; when the defect does not change and the components adjacent to the defect risk points operate normally, a background area wear signal is generated and sent to the optical detection center. After receiving the background area wear signal, the optical detection center adjusts the processing procedure of the PCB board to reduce the wear on the board surface;

[0056] If the image parameters of the defect risk points do not exceed the set image parameter range of the background area, it is inferred that there is a defect risk at the defect risk points. Starting from the current moment, the collected pictures at historical adjacent moments are analyzed. If the image parameters of the defect risk points in the corresponding collected pictures at adjacent moments float numerically in the trend direction of exceeding the image parameter range, it indicates that the defect risk at the defect risk points continues to rise, and a defect warning signal is generated and sent to the optical detection center; after receiving the defect warning signal, the optical detection center repairs the defect risk points and analyzes the collected pictures at historical adjacent moments, traces back to the starting moment when the image parameters of the defect risk points float numerically, and identifies the affected processes based on the processing procedures recorded in the collected pictures at the starting moment and rectifies them in a timely manner;

[0057] If the image parameters of the defect risk points in the corresponding collected pictures at adjacent moments do not float numerically in the trend direction of exceeding the image parameter range, it indicates that the defect risk at the defect risk points is in a controllable state, and a normal wear signal of the board is generated and sent to the optical detection center;

[0058] The optical detection center generates a processing trace analysis signal and sends it to the processing trace analysis unit. After receiving the processing trace analysis signal, the processing trace analysis unit analyzes the processing traces of the mating areas and component points, and infers whether there are defects in the component processing in the PCB that affect the operation feasibility of the entire PCB circuit when the components are processed. By comparing the traces, the accuracy of optical detection can be improved, and further improve the accuracy of defect detection and defect traceability analysis;

[0059] The area where the image parameters of the component points deviate from the image parameters of the background area is marked as the point welding area; where the deviation means the deviation value artificially set by the person in the field during optical detection, which is used to distinguish the component points from the background area;

[0060] Collect the deviation distance between the boundary of the spot welding area at the point and the boundary of the covered area by the spot welding method of the component, where the boundary of the covered area is expressed as the corresponding boundary of the area occupied by the welding slag or the installation of the component during spot welding of the component;

[0061] Collect the boundary of the circuit track and the boundary of the actual circuit welding track within the cooperating area, and obtain the boundary offset of the track according to the comparison of the track boundaries; if there is welding inclination or the expansion of the welding influence area during actual processing, resulting in changes in the track;

[0062] Compare the deviation distance between the boundary of the spot welding area at the point and the boundary of the covered area by the spot welding method of the component, and the boundary offset of the track with the deviation distance value and the offset threshold respectively:

[0063] If the deviation distance between the boundary of the spot welding area at the point and the boundary of the covered area by the spot welding method of the component exceeds the deviation distance value, or the boundary offset of the track exceeds the offset threshold, it is inferred that the analysis of the processing traces is abnormal, generate a processing influence signal and send the processing influence signal to the optical detection center;

[0064] If the deviation distance between the boundary of the spot welding area at the point and the boundary of the covered area by the spot welding method of the component does not exceed the deviation distance value, and the boundary offset of the track does not exceed the offset threshold, it is inferred that the analysis of the processing traces is normal, generate a processing normal signal and send the processing normal signal to the optical detection center;

[0065] After receiving the processing influence signal, the optical detection center generates a processing defect detection signal and sends the processing defect detection signal to the defect recognition and detection unit;

[0066] The defect recognition and detection unit performs defect recognition and detection on the cooperating area and the component points; collect the position of the processing influence and mark the position of the processing influence on the collected picture. If the operation of the component corresponding to the starting moment of the appearance of the processing influence position is not affected, it is marked as deviation processing; and perform change detection on the processing influence position of the deviation processing. If the parameters of the processing influence position in adjacent collected pictures fluctuate, or the operation parameters of the component fluctuate in an abnormal trend, then mark the processing influence position as the position affected by the defect and send it to the optical detection center, and the optical detection center repairs the position affected by the defect; if the parameters of the processing influence position in adjacent collected pictures do not fluctuate, and the operation parameters of the component do not fluctuate in an abnormal trend, then mark the processing influence position as the position not affected by the defect and send it to the optical detection center;

[0067] If the operation of the component corresponding to the starting moment of the appearance of the processing influence position is affected, it is marked as abnormal processing, and adjust the processing welding method and repair the processing influence position at the same time;

[0068] It should be noted that the operation of components is inferred to be normal or abnormal by actual operating parameters such as current, voltage, or signal transmission during operation. An abnormal trend of floating means floating away from the set range of operating parameters; the floating of parameters affecting the machining position is represented by an increase in the depth or length of the defect.

[0069] When the present invention is in use, the lighting impact assessment unit assesses the impact on the lighting module; after the assessment shows no impact, the lighting module is put into use and an image of the PCB board is captured; the image background separation unit processes the captured image, i.e., separates the image background, and obtains the background area, the mutual cooperation area, and the component points through separation; the machining trace analysis unit analyzes the machining traces in the mutual cooperation area and the component points; the defect identification and detection unit identifies and detects defects in the background area, the mutual cooperation area, and the component points.

[0070] The preferred embodiments of the present invention disclosed above are only used to help illustrate the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to the specific implementation manners described. Obviously, many modifications and changes can be made according to the content of this specification. These embodiments are selected and specifically described in this specification to better explain the principle and practical application of the present invention, so that those skilled in the art can understand and utilize the present invention well. The present invention is only limited by the claims and their full scope and equivalents.

Claims

1. An automatic optical inspection device for PCB board defects based on machine vision, characterized in that, It includes an optical detection center, an illumination module, an image acquisition module, and a motion control module; the optical detection center is communicatively connected to: An illumination impact assessment unit that conducts an impact assessment on the illumination module; after the assessment shows no impact, the illumination module is put into use and images of the PCB board are acquired; An image background separation unit that processes the acquired images, i.e., separates the image background, and obtains a background area, a mating area, and component positions through the separation; A processing trace analysis unit that analyzes the processing traces of the mating area and component positions; A defect identification and detection unit that conducts defect identification and detection on the background area, the mating area, and component positions.

2. The automatic optical inspection device for PCB board defects based on machine vision according to claim 1, characterized in that, The process of the illumination impact assessment unit is as follows: When taking any set light source position saved in the illumination module as the real-time light source position, conduct illumination analysis on all components, collect the mean deviation of the marked luminous flux in the near-light area of components at different positions, and at the same time collect the average value of the continuous reduction span of the luminous flux in the far-light area of components at different positions, and conduct a threshold comparison: If the mean deviation of the luminous flux exceeds the luminous flux mean deviation threshold, or the average value of the continuous reduction span of the luminous flux exceeds the continuous reduction span mean threshold, then generate an illumination high-impact signal; if the mean deviation of the luminous flux does not exceed the luminous flux mean deviation threshold, and the average value of the continuous reduction span of the luminous flux does not exceed the continuous reduction span mean threshold, then generate an illumination low-impact signal and send it to the optical detection center and the illumination module.

3. The automatic optical inspection device for PCB board defects based on machine vision according to claim 1, characterized in that, The process of the image background separation unit is as follows: Conduct frame-by-frame picture analysis on the acquired images, identify all the acquired pictures, obtain the on-picture positions of the components inside the PCB through the image ratio, and mark the positions of the components in the acquired pictures; classify the component types into independently operating components and cooperatively operating components according to the working mode type; mark the mating areas of the cooperatively operating components; collect the image parameters of the acquired pictures; mark the points with deviations from the image parameters of the surrounding positions, and compare them with the on-picture positions of the components. If the marked points overlap with the on-picture positions of the components, then mark the surrounding positions of the corresponding marked points as the background area; Correspond the marked points with the on-picture positions of the components one by one, and at the same time collect the mating areas.

4. The automatic optical inspection device for PCB board defects based on machine vision according to claim 3, characterized in that, During the comparison process, if the marked points cannot correspond to the on-picture positions of the corresponding components, then set the corresponding marked points as defect risk points, and compare the image parameters of the defect risk points; If the image parameters of the defect risk points exceed the set image parameter range of the background area, then infer that there are defects at the defect risk points; mark the currently acquired picture as a defect-generated picture, and conduct defect change detection based on adjacent acquired pictures. When the defect changes and the defect area range increases, then generate a background area defect signal; When the defect does not change and the components adjacent to the defect risk points operate normally, then generate a background area wear signal.

5. The automatic optical inspection device for PCB board defects based on machine vision according to claim 4, characterized in that, If the image parameters of the defect risk point do not exceed the set image parameter range of the background area, it is inferred that there is a defect risk at the defect risk point. Starting from the current moment, the captured images at adjacent historical moments are analyzed. If the image parameters of the defect risk point in the captured images at adjacent moments fluctuate numerically in a trend direction that exceeds the image parameter range, a defect warning signal is generated. If the image parameters of the defect risk point in the captured images at adjacent moments do not fluctuate numerically in a trend direction that exceeds the image parameter range, a normal wear signal of the board is generated and sent to the optical inspection center.

6. The automatic optical inspection device for PCB board defects based on machine vision according to claim 1, wherein The process of the processing trace analysis unit is as follows: The area where the image parameters of the component point are deviated from the image parameters of the background area is marked as the point welding area; the deviation distance between the boundary of the point welding area and the covered area of the component spot welding method is collected; the boundary of the circuit track and the actual circuit welding track in the mutually cooperating area are collected, and the boundary offset of the track is obtained by comparing the track boundaries.

7. The automatic optical inspection device for PCB board defects based on machine vision according to claim 6, characterized in that, The deviation distance between the boundary of the point welding area and the covered area of the component spot welding method and the boundary offset of the track are respectively compared with the deviation distance value and the offset threshold: If the deviation distance between the boundary of the point welding area and the covered area of the component spot welding method exceeds the deviation distance value, or the boundary offset of the track exceeds the offset threshold, a processing influence signal is generated; if the deviation distance between the boundary of the point welding area and the covered area of the component spot welding method does not exceed the deviation distance value, and the boundary offset of the track does not exceed the offset threshold, a normal processing signal is generated and sent to the optical inspection center.

8. The automatic optical inspection device for PCB board defects based on machine vision according to claim 7, wherein, The defect identification and detection unit conducts defect identification and detection on the mutually cooperating area and the component points; the position of the processing influence is collected and marked on the captured image. If the operation of the component corresponding to the starting moment of the appearance of the processing influence position is not affected, it is marked as deviation processing; and the change detection of the processing influence position of the deviation processing is carried out; if the operation of the component corresponding to the starting moment of the appearance of the processing influence position is affected, it is marked as abnormal processing, and the processing welding method is adjusted and the processing influence position is repaired at the same time.

9. The automatic optical inspection device for PCB board defects based on machine vision according to claim 8, wherein, The change detection is as follows: If the parameters of the processing influence position in adjacent captured images fluctuate, or the operation parameters of the component fluctuate in an abnormal trend, the processing influence position is marked as the position where the defect has an impact and sent to the optical inspection center, and the optical inspection center repairs the position where the defect has an impact; if the parameters of the processing influence position in adjacent captured images do not fluctuate, and the operation parameters of the component do not fluctuate in an abnormal trend, the processing influence position is marked as the position where the defect has no impact and sent to the optical inspection center.

Images