SMT welding defect identification and detection method and system based on AOI detection

Abstract

The invention relates to the technical field of welding defect detection, in particular to an SMT welding defect identification and detection method and system based on AOI detection, and the system comprises an SMT welding defect identification platform, a detection information unit, a consistency detection unit, a contour detection unit, a multi-angle verification unit and a feedback response unit. According to the method, analysis is carried out from two angles of AOI detection equipment and appearance contour detection, so that the accuracy condition of the detection result of the to-be-detected welding part with unknown defects can be known, and meanwhile, whether the to-be-detected welding part with unknown defects has defects and defect types can be known; the to-be-detected welding part with unknown defects can be reasonably managed in a targeted manner, and the AOI detection equipment is subjected to standard detection and repetitive detection, so that the detection reliability and effectiveness of the AOI detection equipment are ensured, the detection influence of the AOI detection equipment on the to-be-detected welding part with unknown defects is reduced, and the detection efficiency is improved. And interactive analysis is carried out in a progressive manner, so that an output result can be accurately known.

Description

Technical Field

[0001] The present invention relates to the technical field of welding defect detection, and in particular to an SMT welding defect recognition detection method and system based on AOI detection. Background Art

[0002] The principle of AOI inspection is to output the reflected light intensity of the inspected object in a quantitative grayscale value through camera technology, and analyze and determine the defects and classify them by comparing with the grayscale value of the standard image. AOI inspection equipment usually uses a high-definition CCD camera to automatically scan the inspected product, collect images, and then compare the test points with the qualified parameters in the database. After image processing, the defects on the target product are checked and displayed through a display or automatic mark for maintenance personnel to repair and improve the process;

[0003] The main equipment used for SMT welding includes placement machines, reflow machines and hot air workbenches; placement machines are used to accurately place components, reflow machines are used to solder components, and hot air workbenches are suitable for soldering multi-pin components;

[0004] However, in the existing SMT welding defect recognition and detection technology, the AOI detection equipment can only be used for detection in a single way, and the output results of the AOI detection equipment cannot be verified, which reduces the detection credibility of the AOI detection equipment, and cannot be compared and analyzed from the perspective of the outer contour of the welded parts, thereby reducing the defect recognition and detection efficiency, which is not conducive to the subsequent management of the welded parts;

[0005] In view of the above technical defects, a solution is now proposed. Summary of the invention

[0006] The purpose of the present invention is to provide an SMT welding defect recognition and detection method and system based on AOI detection to solve the technical defects mentioned above. The present invention analyzes from two angles: AOI detection equipment and appearance contour detection, which helps to understand the accuracy of the detection results of the unknown defect welded parts to be detected, and at the same time understand whether the unknown defect welded parts to be detected have defects and the defect types, so as to facilitate the reasonable and targeted management of the unknown defect welded parts to be detected. By performing standard detection and repeatability detection on the AOI detection equipment, the detection reliability and effectiveness of the AOI detection equipment are guaranteed to reduce the detection influence of the AOI detection equipment on the unknown defect welded parts to be detected. By performing appearance contour difference supervision feedback analysis on the welding feature image, it is possible to analyze from the perspective of appearance contour to understand whether the unknown defect welded parts to be detected have defects, and provide comparison data for subsequent analysis. Interactive analysis is performed in a progressive manner to accurately understand the output results of the unknown defect welded parts to be detected.

[0007] Preferably, the purpose of the present invention can be achieved by the following technical solutions: an SMT welding defect recognition and detection system based on AOI detection, comprising an SMT welding defect recognition platform, a detection information unit, a consistency detection unit, a contour detection unit, a multi-angle verification unit and a feedback response unit;

[0008] The detection information unit is used to collect the detection report and welding feature image output by the AOI detection equipment;

[0009] The consistency detection unit is used to perform credibility and repeatability verification feedback analysis on the test report output by the AOI detection equipment, perform discrimination analysis on the obtained operation efficiency qualification index, and obtain a standard signal or a deviation signal;

[0010] After the standard signal is generated, the first test report is further subjected to defect discrimination analysis to obtain a defect signal or a normal signal;

[0011] Then, under the condition that the AOI inspection equipment inspects the unknown defect welded part for the first time, the inspection is performed again for g times, where g is a natural number greater than zero, and the obtained consistent inspection index is discriminated and processed to obtain a repeated inspection instruction or a feedback instruction;

[0012] When the feedback instruction is generated, the repeated consistency index is discriminated and analyzed progressively to obtain a normal detection signal or a detection defect signal;

[0013] After receiving the welding feature image, the contour detection unit immediately performs appearance contour difference supervision feedback analysis on the welding feature image to obtain an appearance qualified signal or an appearance unqualified signal;

[0014] When a normal detection signal is generated, the multi-angle verification unit responds to the normal detection signal, performs interactive analysis of the detection information, and sends the obtained consistent signal or re-inspection signal to the feedback response unit.

[0015] Preferably, the credibility repeatability verification feedback analysis process of the consistency detection unit is as follows:

[0016] The operation time period of the welding equipment is collected and set as the time threshold, i welded parts to be inspected with known defects are randomly selected, where i is a natural number greater than zero, and the welded parts to be inspected are placed inside the AOI inspection equipment for inspection, and the inspection report output by the AOI inspection equipment is obtained, and the inspection report output by the AOI inspection equipment is verified with the welded parts to be inspected with known defects, and then the verification standard value of the welded parts to be inspected is obtained, and the verification standard value indicates that the inspection report output by the AOI inspection equipment corresponds to the known defects of the welded parts to be inspected, and the ratio of the verification standard value to i is set as the operation efficiency qualification index, and the operation efficiency qualification index is discriminated and analyzed to obtain a standard signal or a deviation signal.

[0017] Preferably, after the standard signal is generated, the welded parts to be detected with unknown defects are obtained, the welded parts to be detected with unknown defects are detected by an AOI detection device, the first detection report output by the AOI detection device is obtained, and defect discrimination analysis is performed on the first detection report:

[0018] If the first test report shows a defect, a defect signal is generated;

[0019] If the first test report shows no defects, a normal signal is generated.

[0020] Preferably, the AOI detection equipment then performs another detection g times under the condition of the first detection of the welded parts to be detected with unknown defects, where g is a natural number greater than zero, and each detection time period of the AOI detection equipment under the condition of the first detection within the time threshold is obtained, and set as the analysis time period, and the operating parameters and lighting parameters of the AOI detection equipment within the analysis time period are obtained, the operating parameters include the operating voltage and the operating temperature, and the lighting parameters include the lighting angle and the lighting intensity, and the operating parameters and lighting parameters of the AOI detection equipment in each sub-time period are discriminated and processed, if the operating parameters and lighting parameters are the same as the corresponding preset operating parameters and preset lighting parameters, then the analysis time period is determined to be a normal time period, the number corresponding to the normal time period is obtained, and the number corresponding to the normal time period is set as the consistent detection index, and the consistent detection index is discriminated and processed to obtain a repeated detection instruction or a feedback instruction.

[0021] Preferably, when generating a feedback instruction:

[0022] The inspection report output by the AOI inspection device for each inspection number corresponding to the feedback instruction is obtained, and the inspection report output by the AOI inspection device for each inspection number corresponding to the feedback instruction is compared and analyzed with the reference report, and the number of difference inspections between the inspection report output by the AOI inspection device for each inspection number corresponding to the feedback instruction and the reference report is obtained, and it is set as the repetition consistency index, and the repetition consistency index is discriminated and analyzed to obtain a normal inspection signal or a defect inspection signal.

[0023] Preferably, the appearance contour difference supervision feedback analysis process of the contour detection unit is as follows:

[0024] A welding feature image of a welded part to be detected with unknown defects within a time threshold is obtained, welding information at the welding point in the welding feature image is obtained, the welding information includes a welding feature contour line and a solder evaluation index, a difference value between the welding feature contour line and a standard welding feature contour line is obtained, and the difference value between the welding feature contour line and the standard welding feature contour line is set as a contour deviation value, the contour deviation value and the solder evaluation index are compared and analyzed with a preset contour deviation value range and a preset solder evaluation index threshold that are internally entered and stored, and an appearance qualified signal or an appearance unqualified signal is obtained.

[0025] The welding characteristic contour line represents the outer contour line of the characteristic shape of the solder surface at the welding point; the solder evaluation index represents the number of deviations between the solder paste amount at the welding point and the actual welding contact area from the preset corresponding threshold value.

[0026] Preferably, the interactive analysis process of the detection information of the multi-angle verification unit is as follows:

[0027] Appearance qualified signals, appearance unqualified signals, normal signals and defect signals are obtained, and the appearance qualified signals, appearance unqualified signals, normal signals and defect signals are analyzed to obtain consistent signals or re-inspection signals.

[0028] The beneficial effects of the present invention are as follows:

[0029] (1) The present invention analyzes from two perspectives: AOI detection equipment and appearance profile detection, thereby helping to understand the accuracy of the detection results of the welded parts to be detected with unknown defects, and at the same time understand whether the welded parts to be detected with unknown defects have defects and the types of defects, so as to facilitate reasonable and targeted management of the welded parts to be detected with unknown defects, and by performing standard detection and repeatability detection on the AOI detection equipment, the detection reliability and effectiveness of the AOI detection equipment are guaranteed, so as to reduce the detection impact of the AOI detection equipment on the welded parts to be detected with unknown defects;

[0030] (2) The present invention performs appearance contour difference supervision feedback analysis on the welding feature image so as to analyze from the perspective of appearance contour to understand whether the welded parts to be inspected with unknown defects have defects, and provide comparison data for subsequent analysis, and interactive analysis is performed in a progressive manner to accurately understand the output results of the welded parts to be inspected with unknown defects. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] The present invention will be further described below in conjunction with the accompanying drawings;

[0032] Figure 1 It is a flowchart of the system of the present invention;

[0033] Figure 2 It is a reference diagram of the method of the present invention. DETAILED DESCRIPTION

[0034] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0035] Embodiment 1:

[0036] See also Figure 1 to Figure 2 As shown, the present invention is an SMT welding defect recognition and detection system based on AOI detection, including an SMT welding defect recognition platform, a detection information unit, a consistency detection unit, a contour detection unit, a multi-angle verification unit and a feedback response unit, the SMT welding defect recognition platform is connected to the detection information unit in a one-way communication, the detection information unit is connected to the consistency detection unit and the contour detection unit in a one-way communication, the consistency detection unit is connected to the multi-angle verification unit and the feedback response unit in a one-way communication, the contour detection unit is connected to the multi-angle verification unit in a one-way communication, and the multi-angle verification unit is connected to the feedback response unit in a one-way communication;

[0037] When the SMT welding defect recognition platform generates a supervision instruction, it sends the supervision instruction to the detection information unit. After receiving the supervision instruction, the detection information unit immediately collects the test report and welding feature image output by the AOI detection equipment, and sends the test report and welding feature image output by the AOI detection equipment to the consistency detection unit and the contour detection unit respectively. After receiving the test report output by the AOI detection equipment, the consistency detection unit immediately performs credibility repeatability verification feedback analysis on the output test report. On the one hand, it determines the credibility of the test results of the AOI detection equipment, and on the other hand, it helps to reasonably and targetedly manage the unknown defect welded parts to be detected. The specific credibility repeatability verification feedback analysis process is as follows:

[0038] The operation time period of the welding equipment is collected and set as the time threshold. i welded parts to be inspected with known defects are randomly selected, where i is a natural number greater than zero. The welded parts to be inspected are placed inside the AOI inspection equipment for inspection, and the inspection report output by the AOI inspection equipment is obtained. The inspection report output by the AOI inspection equipment is verified with the welded parts to be inspected with known defects, and then the verification standard value of the welded parts to be inspected is obtained. The verification standard value indicates that the inspection report output by the AOI inspection equipment corresponds to the known defects of the welded parts to be inspected, and the ratio of the verification standard value to i is set as the qualified operation efficiency index, and the qualified operation efficiency index is discriminated and analyzed:

[0039] If the operational efficiency qualification index is equal to 1, a standard signal is generated;

[0040] If the efficiency qualification index is not equal to 1, a deviation signal is generated and sent to the feedback response unit. After receiving the deviation signal, the feedback response unit immediately displays the preset warning text corresponding to the deviation signal, so as to adjust the AOI detection equipment to ensure the accuracy of the output result of the AOI detection equipment;

[0041] In the embodiment of the present invention, known defects include bridging, tin overflow, and non-printing;

[0042] After the standard signal is generated, the welded parts to be inspected with unknown defects are obtained, and the welded parts to be inspected with unknown defects are inspected by the AOI inspection equipment, and the first inspection report output by the AOI inspection equipment is obtained, and the defect discrimination analysis is performed on the first inspection report:

[0043] If the first test report shows a defect, a defect signal is generated;

[0044] If the first test report shows no defects, a normal signal is generated;

[0045] Get the first inspection report output by the AOI inspection equipment and set it as a reference report;

[0046] Then, under the condition that the AOI detection equipment detects the unknown defect welded parts for the first time, it detects again g times, where g is a natural number greater than zero, and obtains each detection period of the AOI detection equipment under the condition of the first detection within the time threshold, and sets it as the analysis period, obtains the operating parameters and illumination parameters of the AOI detection equipment within the analysis period, the operating parameters include operating voltage, operating temperature, etc., and the illumination parameters include illumination angle, illumination intensity, etc., and the operating parameters and illumination parameters of the AOI detection equipment in each sub-time period are discriminated and processed. If the operating parameters and illumination parameters are the same as the corresponding preset operating parameters and preset illumination parameters, then the analysis period is determined to be a normal period, the number corresponding to the normal period is obtained, and the number corresponding to the normal period is set as the consistent detection index, and the consistent detection index is discriminated and processed:

[0047] If the consistent detection index is less than the preset consistent detection index threshold, a repeat detection instruction is generated. When the repeat detection instruction is given, the welded part to be detected with unknown defects is tested again under the conditions of the first test until a feedback instruction is generated;

[0048] If the consistent detection index is greater than or equal to the preset consistent detection index threshold, a feedback instruction is generated. When the feedback instruction is generated, the detection report output by the AOI detection device for each detection number corresponding to the feedback instruction is obtained, and the detection report output by the AOI detection device for each detection number corresponding to the feedback instruction is compared and analyzed with the reference report, and the number of different detections between the detection report output by the AOI detection device for each detection number corresponding to the feedback instruction and the reference report is obtained, and it is set as the repeated consistency index, and the repeated consistency index is discriminated and analyzed:

[0049] If the repeat consistency index is equal to zero, a detection normal signal is generated;

[0050] If the repetition consistency index is not equal to zero, a detection defect signal is generated, and the detection normal signal and the detection defect signal are sent to the feedback response unit. After receiving the detection normal signal and the detection defect signal, the feedback response unit immediately displays the preset warning text corresponding to the detection normal signal and the detection defect signal to understand whether the multiple detection output reports of the AOI detection equipment under the same conditions are the same, so as to understand the credibility and accuracy of the output results of the AOI detection equipment.

[0051] Embodiment 2:

[0052] After receiving the welding feature image, the contour detection unit immediately performs appearance contour difference supervision feedback analysis on the welding feature image, so as to analyze it from the perspective of appearance contour to understand whether the welded part to be detected with unknown defects has defects. The specific appearance contour difference supervision feedback analysis process is as follows:

[0053] The welding feature image of the welded part to be detected with unknown defects within the time threshold is obtained, the welding information at the welding point in the welding feature image is obtained, the welding information includes the welding feature contour line and the solder evaluation index, the difference value between the welding feature contour line and the standard welding feature contour line is obtained, and the difference value between the welding feature contour line and the standard welding feature contour line is set as the contour deviation value, and the contour deviation value and the solder evaluation index are compared and analyzed with the preset contour deviation value range and the preset solder evaluation index threshold value recorded and stored internally:

[0054] If the contour deviation value belongs to the preset contour deviation value range, and the solder evaluation index is equal to the preset solder evaluation index threshold, a qualified appearance signal is generated;

[0055] If the contour deviation value does not belong to the preset contour deviation value range, or the solder evaluation index is not equal to the preset solder evaluation index threshold, an appearance failure signal is generated;

[0056] In the embodiment of the present invention, the welding characteristic contour line represents the contour line of the characteristic shape of the solder surface layer at the welding point. It should be noted that the SMT welding defects are identified and detected from the perspective of the appearance to ensure the accuracy of the detection result;

[0057] In the embodiment of the present invention, the solder evaluation index indicates the number of deviations of the solder paste amount and the actual contact area of ​​the soldering point from the preset corresponding threshold value, and is analyzed from the perspective of actual soldering to understand whether the SMT soldering is qualified;

[0058] When a normal detection signal is generated, the multi-angle verification unit responds to the normal detection signal and performs interactive analysis of the detection information, so as to rationally manage the welded parts to be detected with unknown defects. The specific interactive analysis process of the detection information is as follows:

[0059] Obtain appearance qualified signals, appearance unqualified signals, normal signals, and defective signals, and analyze the appearance qualified signals, appearance unqualified signals, normal signals, and defective signals:

[0060] If a qualified appearance signal and a normal signal are generated, or a unqualified appearance signal and a defect signal are generated, a consistent signal is obtained;

[0061] If an appearance qualified signal and a defect signal, or an appearance unqualified signal and a normal signal are generated, a re-inspection signal is generated, and the consistent signal and the re-inspection signal are sent to the feedback response unit. After receiving the consistent signal and the re-inspection signal, the feedback response unit immediately displays the preset warning text corresponding to the consistent signal and the re-inspection signal, that is, the preset warning text corresponding to the consistent signal is "the detection result of the AOI detection equipment", and the preset warning text corresponding to the re-inspection signal is "manual verification", so as to intuitively understand the detection results of the welded parts to be detected with unknown defects, and at the same time facilitate manual re-inspection of the welded parts to be detected with controversial unknown defects, which helps to improve the detection accuracy of the welded parts to be detected with unknown defects, and at the same time facilitates reasonable and targeted management of the welded parts to be detected with unknown defects.

[0062] Embodiment three:

[0063] The SMT welding defect recognition and detection method based on AOI detection includes the following steps:

[0064] Step 1: Perform standard inspection on SMT welded parts with known defects to determine whether the inspection report output by the AOI inspection equipment is accurate. If a standard signal is generated, proceed to step 2. If a deviation signal is generated, feedback management is performed.

[0065] Step 2: The first detection of the welded part to be detected with unknown defects under the standard signal is performed to obtain a defect signal or a normal signal;

[0066] Step 3: Based on the repeated inspection analysis under the first inspection conditions of the AOI inspection equipment, the obtained consistent inspection index is judged and processed. If a repeated inspection instruction is obtained, feedback management is performed. If a feedback instruction is obtained, step 4 is entered;

[0067] Step 4: Based on the comparison and analysis of the test reports output by the AOI test equipment at each test time and the reference report, the obtained repetitive consistency index is judged and processed. If a normal test signal is obtained, step 6 is entered. If a test defect signal is obtained, feedback management is performed;

[0068] Step 5: Collect the welding feature image of the welded part to be inspected with unknown defects to conduct supervision feedback analysis on the difference of appearance contours, and obtain a qualified appearance signal or a failed appearance signal;

[0069] Step 6: Interactively analyze the inspection information of the qualified appearance signal, the unqualified appearance signal, the normal signal, and the defect signal in a progressive manner to obtain a consistent signal or a re-inspection signal;

[0070] In summary, the present invention performs analysis from two perspectives, AOI detection equipment and appearance profile detection, which helps to understand the accuracy of the detection results of the welded parts to be detected with unknown defects, and at the same time understand whether the welded parts to be detected with unknown defects have defects and the types of defects, so as to facilitate reasonable and targeted management of the welded parts to be detected with unknown defects, and perform standard detection and repeatability detection on the AOI detection equipment to ensure the detection reliability and effectiveness of the AOI detection equipment, so as to reduce the detection impact of the AOI detection equipment on the welded parts to be detected with unknown defects, and perform appearance profile difference supervision feedback analysis on the welding feature image, so as to analyze from the perspective of appearance profile, so as to understand whether the welded parts to be detected with unknown defects have defects, and provide comparison data for subsequent analysis, and perform interactive analysis in a progressive manner to accurately understand the output results of the welded parts to be detected with unknown defects.

[0071] The threshold is set to facilitate comparison. The size of the threshold depends on the amount of sample data and the number of bases set by technicians in this field for each group of sample data; as long as it does not affect the proportional relationship between the parameter and the quantized value.

[0072] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. The SMT welding defect recognition and detection system based on AOI detection is characterized by: It includes SMT welding defect recognition platform, detection information unit, consistency detection unit, contour detection unit, multi-angle verification unit and feedback response unit; The detection information unit is used to collect the detection report and welding feature image output by the AOI detection equipment; The consistency detection unit is used to perform credibility and repeatability verification feedback analysis on the test report output by the AOI detection equipment, perform discrimination analysis on the obtained operation efficiency qualification index, and obtain a standard signal or a deviation signal; After the standard signal is generated, the first test report is further subjected to defect discrimination analysis to obtain a defect signal or a normal signal; Then, under the condition that the AOI inspection equipment inspects the unknown defect welded part for the first time, the inspection is performed again for g times, where g is a natural number greater than zero, and the obtained consistent inspection index is discriminated and processed to obtain a repeated inspection instruction or a feedback instruction; When the feedback instruction is generated, the repeated consistency index is discriminated and analyzed progressively to obtain a normal detection signal or a detection defect signal; After receiving the welding feature image, the contour detection unit immediately performs appearance contour difference supervision feedback analysis on the welding feature image to obtain an appearance qualified signal or an appearance unqualified signal; When a normal detection signal is generated, the multi-angle verification unit responds to the normal detection signal, performs interactive analysis of the detection information, and sends the obtained consistent signal or re-inspection signal to the feedback response unit.

2. The SMT welding defect recognition and detection system based on AOI detection according to claim 1 is characterized in that: The credibility repeatability verification feedback analysis process of the consistency detection unit is as follows: The operation time period of the welding equipment is collected and set as the time threshold, i welded parts to be inspected with known defects are randomly selected, where i is a natural number greater than zero, and the welded parts to be inspected are placed inside the AOI inspection equipment for inspection, and the inspection report output by the AOI inspection equipment is obtained, and the inspection report output by the AOI inspection equipment is verified with the welded parts to be inspected with known defects, and then the verification standard value of the welded parts to be inspected is obtained, and the verification standard value indicates that the inspection report output by the AOI inspection equipment corresponds to the known defects of the welded parts to be inspected, and the ratio of the verification standard value to i is set as the operation efficiency qualification index, and the operation efficiency qualification index is discriminated and analyzed to obtain a standard signal or a deviation signal.

3. The SMT welding defect recognition and detection system based on AOI detection according to claim 2 is characterized in that: After the standard signal is generated, the welded parts to be inspected with unknown defects are obtained, and the welded parts to be inspected with unknown defects are inspected by the AOI inspection equipment, and the first inspection report output by the AOI inspection equipment is obtained, and the defect discrimination analysis is performed on the first inspection report: If the first test report shows a defect, a defect signal is generated; If the first test report shows no defects, a normal signal is generated.

4. The SMT welding defect recognition and detection system based on AOI detection according to claim 3 is characterized in that: Then, under the condition that the AOI detection equipment detects the unknown defect welding piece to be detected for the first time, it detects again g times, g is a natural number greater than zero, and each detection time period of the AOI detection equipment under the condition of the first detection within the time threshold is obtained, and it is set as the analysis time period, and the operating parameters and lighting parameters of the AOI detection equipment in the analysis time period are obtained, the operating parameters include operating voltage and operating temperature, and the lighting parameters include lighting angle and lighting intensity, and the operating parameters and lighting parameters of the AOI detection equipment in each sub-time period are discriminated and processed. If the operating parameters and lighting parameters are the same as the corresponding preset operating parameters and preset lighting parameters, then the analysis time period is determined to be a normal time period, and the number corresponding to the normal time period is obtained, and the number corresponding to the normal time period is set as a consistent detection index, and the consistent detection index is discriminated and processed to obtain a repeated detection instruction or a feedback instruction.

5. The SMT welding defect recognition and detection system based on AOI detection according to claim 4 is characterized in that: When generating a feedback instruction: The inspection report output by the AOI inspection device for each inspection number corresponding to the feedback instruction is obtained, and the inspection report output by the AOI inspection device for each inspection number corresponding to the feedback instruction is compared and analyzed with the reference report, and the number of difference inspections between the inspection report output by the AOI inspection device for each inspection number corresponding to the feedback instruction and the reference report is obtained, and it is set as the repetition consistency index, and the repetition consistency index is discriminated and analyzed to obtain a normal inspection signal or a defect inspection signal.

6. The SMT welding defect recognition and detection system based on AOI detection according to claim 5 is characterized in that: The appearance contour difference supervision feedback analysis process of the contour detection unit is as follows: A welding feature image of a welded part to be detected with unknown defects within a time threshold is obtained, welding information at the welding point in the welding feature image is obtained, the welding information includes a welding feature contour line and a solder evaluation index, a difference value between the welding feature contour line and a standard welding feature contour line is obtained, and the difference value between the welding feature contour line and the standard welding feature contour line is set as a contour deviation value, the contour deviation value and the solder evaluation index are compared and analyzed with a preset contour deviation value range and a preset solder evaluation index threshold that are internally entered and stored, and an appearance qualified signal or an appearance unqualified signal is obtained. The welding characteristic contour line represents the outer contour line of the characteristic shape of the solder surface at the welding point; the solder evaluation index represents the number of deviations between the solder paste amount at the welding point and the actual welding contact area from the preset corresponding threshold value.

7. The SMT welding defect recognition and detection system based on AOI detection according to claim 6 is characterized in that: The interactive analysis process of the detection information of the multi-angle verification unit is as follows: Appearance qualified signals, appearance unqualified signals, normal signals and defect signals are obtained, and the appearance qualified signals, appearance unqualified signals, normal signals and defect signals are analyzed to obtain consistent signals or re-inspection signals.

8. An SMT welding defect recognition and detection method based on AOI detection, the method being applied to the SMT welding defect recognition and detection system based on AOI detection as claimed in claim 7, characterized in that: The following steps are involved: Step 1: Perform standard inspection on SMT welded parts with known defects to determine whether the inspection report output by the AOI inspection equipment is accurate. If a standard signal is generated, proceed to step 2. If a deviation signal is generated, feedback management is performed. Step 2: The first detection of the welded part to be detected with unknown defects under the standard signal is performed to obtain a defect signal or a normal signal; Step 3: Based on the repeated inspection analysis under the first inspection conditions of the AOI inspection equipment, the obtained consistent inspection index is judged and processed. If a repeated inspection instruction is obtained, feedback management is performed. If a feedback instruction is obtained, step 4 is entered; Step 4: Based on the comparison and analysis of the test reports output by the AOI test equipment at each test time and the reference report, the obtained repetitive consistency index is judged and processed. If a normal test signal is obtained, step 6 is entered. If a test defect signal is obtained, feedback management is performed; Step 5: Collect the welding feature image of the welded part to be inspected with unknown defects to conduct supervision feedback analysis on the difference of appearance contours, and obtain a qualified appearance signal or a failed appearance signal; Step 6: Based on a progressive approach, interactively analyze the detection information of the appearance qualified signal, appearance unqualified signal, normal signal and defect signal to obtain a consistent signal or a re-inspection signal.

Images