A method and system for evaluating reliability of electronic products

Abstract

The present invention discloses a method and system for evaluating the reliability of electronic products, which relate to the field of electronic product testing. The method and system include: an acquisition module for acquiring circuit board image data, converting the circuit board image data into a grayscale image, and extracting a circuit board contour image from the grayscale image; an upload module for uploading circuit board structural parameters and circuit and component distribution information, and constructing a virtual circuit board three-dimensional model based on the circuit board structural parameters and circuit and component distribution information; the present invention combines image acquisition with three-dimensional modeling to construct a comparative analysis system between the physical image of the circuit board and the virtual model, can improve image clarity based on a grayscale image optimization algorithm, realize accurate digital mapping of the circuit board structure and component distribution through multi-dimensional parameter modeling, introduce an attention level evaluation model, calculate regional attention priority, and combine frequency domain analysis with a pixel difference algorithm to quantify reliability indicators, so that the evaluation results are more in line with the actual defect risk.

Description

Technical Field

[0001] The present invention relates to the technical field of electronic product testing, and in particular to an electronic product reliability assessment method and system. Background Art

[0002] In the reliability assessment of electronic products, appearance inspection is a basic and critical link. For circuit boards, the key to its inspection lies in the accuracy of key positions such as circuits, components, and solder joints on the circuit boards, thereby ensuring that the appearance quality of the circuit boards meets the design and use requirements.

[0003] The invention patent application with application number 202310238796.3 discloses a method and system for reliability verification of electronic products, which method includes at least the following steps: Step A: loading a preset first load on the boundary of the whole-machine simulation model of the electronic product to obtain the response result of the boundary of the focus part in the electronic product to the first load in the whole-machine simulation model, and extracting the first simulation verification characteristic parameter of the focus part in the whole-machine simulation model; Step B: extracting the structural model of the focus part from the whole-machine structure model of the electronic product as a local structural model, and setting the boundary of the local structural model according to the boundary of the focus part in the whole-machine structure model; Step C: establishing a simulation model based on the local structural model after setting the boundary as a local simulation model; Step D: loading a preset second load on the boundary of the local simulation model, and extracting the second simulation verification characteristic parameter of the focus part in the local simulation model, the second load is set in the whole-machine simulation according to the boundary of the focus part. The response result of the first load is set in the model: the boundary of the local simulation model is corrected according to the simulation verification characteristic parameters extracted twice; step E: the boundary of the local physical sample corresponding to the part of interest is set to be consistent with the boundary of the local simulation model after correction; step F: the local physical sample is tested according to the second load loaded by the boundary of the local simulation model to verify the reliability of the part of interest. This application aims to solve the problem that "the reliability of the internal structure and components of electronic products must be verified through a complete whole machine. Due to sealing considerations, it is impossible to arrange acceleration and temperature sensors and other monitoring processes inside, and the accuracy of determining the failure moment of a certain internal position is poor; after the whole machine test, the mechanical structure and other internal components cannot be reused due to the accumulated damage, resulting in a waste of resources and high material cost for a single test; for large-volume and heavy-weight whole machine thermal-vibration tests, a large-tonnage physical vibration table with an ambient temperature chamber is required, the test cost is high, and the heating and cooling rate is slow, which occupies resources for a long time."

[0004] However, for circuit boards, the existing technology uses visual inspection to globally perform appearance inspection on the distribution posture and shape of circuits and components on the circuit boards. The inspection effect on key areas on the circuit boards is weak. During the inspection process, the proportion of each key area on the circuit board is not mapped into the inspection results.

[0005] To this end, we propose an electronic product reliability assessment method and system. Summary of the Invention

[0006] In view of the above-mentioned shortcomings of the prior art, the present invention provides an electronic product reliability assessment method and system, which can effectively solve the problems of the prior art.

[0007] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0008] The present invention discloses an electronic product reliability evaluation system, comprising:

[0009] An acquisition module is used to acquire circuit board image data, convert the circuit board image data into a grayscale image, and extract a circuit board contour image from the grayscale image; an upload module is used to upload circuit board structural parameters and circuit and component distribution information, and construct a virtual circuit board three-dimensional model based on the circuit board structural parameters and circuit and component distribution information; an analysis module is used to receive the virtual circuit board three-dimensional model and the circuit board contour image, collect a comparison analysis image group from the virtual circuit board three-dimensional model and the circuit board contour image, and analyze the circuit board reliability index based on the comparison analysis image group; an evaluation module is used to receive the analysis results of each circuit board reliability index in the analysis module, and evaluate the circuit board reliability based on each analysis result; a judgment module is used to obtain the circuit board reliability evaluation result in the evaluation module, set a circuit board qualification judgment threshold, compare the judgment threshold with the circuit board reliability evaluation result, and judge that the circuit board is qualified when the circuit board reliability evaluation result is greater than or equal to the judgment threshold; a recording module is used to trigger operation when the judgment module judges that the circuit board is unqualified, and record information related to the defects of the unqualified circuit board.

[0010] Furthermore, the circuit board image data collected by the acquisition module is a plane image of one side of the circuit board and the distribution of components on the circuit board. After obtaining the circuit board contour image, the acquisition module stores the circuit board contour image.

[0011] After the acquisition module acquires the circuit board image data and converts the circuit board image data into a grayscale image, it first optimizes the grayscale image and then performs an extraction operation;

[0012] The optimization processing logic of the grayscale image is expressed as:

[0013] ;

[0014] Where: This is the grayscale image of the circuit board after optimization; is the original circuit board grayscale image; is the global mean of the image; is the global standard deviation of the image; is the image gradient; Mask the region of interest; is a hyperparameter;

[0015] Among them, the value ranges of each hyperparameter are , They are used to adjust the scaling ratio of the image grayscale value, control the overall offset of the image grayscale value, determine the degree of edge enhancement, and adjust the contrast of the ROI area.

[0016] Furthermore, when constructing the virtual PCB three-dimensional model, the upload module applies the PCB structural parameters to construct a carrier model of the circuits and components on the PCB, and further constructs a circuit and component distribution model on the PCB based on the circuit and component distribution information on the carrier model. The combination of the carrier module and the circuit and component distribution model on the PCB is recorded as the virtual PCB three-dimensional model.

[0017] Among them, the solder joints of each circuit and component on the circuit board are all represented on the virtual circuit board three-dimensional model.

[0018] Furthermore, the upload module is provided with a selection unit and a marking unit at a lower level. The selection unit is used to select a marking area on the side where the circuit and component distribution model is located on the virtual circuit board three-dimensional model. The marking unit is used to obtain the marking area selected by the selection module, mark and select the marking area with an image frame, and configure the marking information.

[0019] Among them, during the running stage of the annotation unit, image frame annotation is performed from the perspective of the side where the circuit and component distribution model is located on the three-dimensional model of the virtual circuit board. The size of the image frame is customized by the system user. After the image frame annotation of the annotation area is completed, the configured annotation information is the attention level of the annotation area.

[0020] Furthermore, the attention level of each marked area is subject to:

[0021] ;

[0022] Where: is the attention level of the i-th marked area; is the circuit trace density in the i-th marked area; The global circuit trace density of the source surface image of the marked area; is the importance weight of component j; is whether component j exists in the sub-image; is the adjustment coefficient; is the number of vias in the i-th marked area; is the number of intersections of the lines in the i-th marked area; The total number of global vias and intersections in the source surface image of the marked area; is a constant;

[0023] in, The larger the value, the more attention the marked area deserves. Based on the above formula, the attention level of each marked area is calculated. The importance weight of each component is preset by the system end user. The value is 0 or 1, and half or more of the image area of component j is in the marked area. =1, otherwise, =0, adjustment coefficient The initial value is set to 0.5, and the constant The value follows the rule that the larger the marked area, the smaller the value, and vice versa. >0.

[0024] Furthermore, the source of the comparative analysis image group collected by the analysis module is the marked area obtained by the marking unit and the regional contour image obtained by corresponding segmentation of the marked area in the circuit board contour image;

[0025] Each of the comparison and analysis image groups includes a marked area and an area contour image, and the marked area and the area contour image included in the same comparison and analysis image group correspond to the same area on the circuit board;

[0026] The circuit board reliability index analyzed by the analysis module based on the comparison analysis image group is:

[0027] ;

[0028] Where: It is the reliability index of the circuit board; Annotated area image The gradient amplitude of is the region contour image; It is an adjustment parameter with a value range of 0.01~0.1; is the binary version of the labeled area image; Indicates finding the sum of squares of pixel differences within the brackets; is the total area of the labeled region image; Annotated area image Frequency domain representation of ; is the region contour image Frequency domain representation of ; is a very small positive number;

[0029] in, The larger it is, the more reliable the area contour image participating in the calculation corresponds to the area on the circuit board.

[0030] Furthermore, when evaluating the reliability of the circuit board based on the reliability index analysis results of each circuit board, the evaluation module introduces the attention level of the marked area corresponding to each reliability index analysis result into the evaluation process:

[0031] ;

[0032] Where: is the reliability performance value of the circuit board; The total amount of image groups was analyzed for control; is the control analysis image group obtained corresponding to the vth control analysis image group; for The attention level of the corresponding marked area; is the total amount of the marked area; is the attention level of the i-th marked area.

[0033] Furthermore, during the operation stage of the recording module, the reliability indicators of each focus area corresponding to the circuit board judged as unqualified are obtained, and each reliability indicator is further compared with the judgment threshold in the judgment module, and the focus area where the reliability indicator is less than the judgment threshold is recorded as the record content, that is, the operation of recording the relevant information of the defects of the unqualified circuit board is performed.

[0034] Furthermore, the acquisition module is interactively connected to the upload module through a wireless network, the upload module is interactively connected to the selection unit and the labeling unit through a wireless network, the upload module is interactively connected to the analysis module and the evaluation module through a wireless network, the evaluation module is interactively connected to the determination module and the recording module through a wireless network, and the recording module is interactively connected to the labeling unit through a wireless network.

[0035] In another aspect, a reliability assessment method for an electronic product includes:

[0036] Collect circuit board image data, convert the circuit board image data into a grayscale image, optimize the grayscale image, and extract the circuit board contour image from the optimized grayscale image; upload the circuit board structural parameters and circuit and component distribution information, and build a virtual circuit board three-dimensional model based on the circuit board structural parameters and circuit and component distribution information; obtain a marked area image on the surface of the virtual circuit board three-dimensional model, obtain a regional contour image by comparing the marked area segmentation in the circuit board contour image, and combine the marked area image with the regional contour image to obtain a comparison analysis image group; analyze the circuit board reliability index based on each comparison analysis image group; evaluate the circuit board reliability based on the circuit board reliability index analysis results of each comparison analysis image group; set a judgment threshold, obtain the circuit board reliability evaluation result, and judge whether the circuit board is qualified based on the comparison of the evaluation result and the judgment threshold; identify and record the defective area of the unqualified circuit board.

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

[0038] The present invention provides an electronic product reliability assessment method and system. The method and system construct a comparative analysis system between the physical image and virtual model of a circuit board by combining image acquisition with three-dimensional modeling. The method and system can improve image clarity based on a grayscale image optimization algorithm, and achieve accurate digital mapping of the circuit board structure and component distribution through multi-dimensional parameter modeling. In addition, an attention level assessment model is introduced to dynamically calculate the regional attention priority based on parameters such as circuit density, component importance, and the number of vias. The reliability index is quantified by combining frequency domain analysis and pixel difference algorithm to make the assessment results more consistent with the actual defect risk. The regional attention level weight is integrated into the assessment process to achieve hierarchical reference judgment of key areas. At the same time, through the indicator backtracking mechanism of unqualified circuit boards, the source area of reliability parameters below the threshold is accurately located, providing data support for defect analysis, which is different from traditional single image analysis technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. Those skilled in the art can also derive other drawings based on these drawings without inventive effort.

[0040] Figure 1 The figure is a schematic diagram of the structure of an electronic product reliability evaluation system;

[0041] Figure 2 The figure is a flow chart of a reliability evaluation method for electronic products. DETAILED DESCRIPTION

[0042] To make the purpose, technical solutions, and advantages of the embodiments of the present invention more clear, 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 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 making creative efforts are within the scope of protection of the present invention.

[0043] The present invention will be further described below with reference to the embodiments.

[0044] Example 1:

[0045] An electronic product reliability evaluation system of this embodiment, such as Figure 1 Shown, including:

[0046] An acquisition module is used to acquire circuit board image data, convert the circuit board image data into a grayscale image, and extract a circuit board contour image from the grayscale image;

[0047] The circuit board image data collected by the acquisition module is a plane image of one side of the circuit board and the distribution of components. After obtaining the circuit board contour image, the acquisition module stores the circuit board contour image;

[0048] After the acquisition module acquires the circuit board image data and converts the circuit board image data into a grayscale image, it first optimizes the grayscale image and then performs the extraction operation;

[0049] The optimization processing logic of grayscale images is expressed as:

[0050] ;

[0051] Where: This is the grayscale image of the circuit board after optimization; is the original circuit board grayscale image; is the global mean of the image; is the global standard deviation of the image; is the image gradient; Mask the region of interest; is a hyperparameter;

[0052] Among them, the value ranges of each hyperparameter are , They are used to adjust the scaling ratio of the image grayscale value, control the overall offset of the image grayscale value, determine the degree of edge enhancement, and adjust the contrast of the ROI area;

[0053] Through the above logic formula, the grayscale image of the circuit board is optimized to improve the accuracy of the system operation processing data, thereby improving the accuracy of the system operation output results;

[0054] An upload module is used to upload circuit board structural parameters and circuit and component distribution information, and to build a virtual circuit board three-dimensional model based on the circuit board structural parameters and circuit and component distribution information;

[0055] When constructing the virtual PCB three-dimensional model, the upload module applies the PCB structural parameters to construct a carrier model of the circuits and components on the PCB, and further constructs a circuit and component distribution model on the PCB based on the circuit and component distribution information on the carrier model. The combination of the carrier module and the circuit and component distribution model on the PCB is recorded as the virtual PCB three-dimensional model.

[0056] Among them, the solder joints of each circuit and component on the circuit board are represented on the virtual circuit board three-dimensional model;

[0057] The upload module is provided with a selection unit and a marking unit at the lower level. The selection unit is used to select a marking area on the side where the circuit and component distribution model are located on the virtual circuit board 3D model. The marking unit is used to obtain the marking area selected in the selection module, mark and select the marking area with an image frame, and configure the marking information.

[0058] During the annotation unit operation phase, image frame annotation is performed from the perspective of the side where the circuit and component distribution model is located on the three-dimensional model of the virtual circuit board. The size of the image frame is customized by the system user. After the image frame annotation is completed, the annotation information configured is the attention level of the annotation area.

[0059] The attention level of each marked area is subject to:

[0060] ;

[0061] Where: is the attention level of the i-th marked area; is the circuit trace density in the i-th marked area; The global circuit trace density of the source surface image of the marked area; is the importance weight of component j; is whether component j exists in the sub-image; is the adjustment coefficient; is the number of vias in the i-th marked area; is the number of intersections of the lines in the i-th marked area; The total number of global vias and intersections in the source surface image of the marked area; is a constant;

[0062] in, The larger the value, the more attention the marked area deserves. Based on the above formula, the attention level of each marked area is calculated. The importance weight of each component is preset by the system end user. The value is 0 or 1, and half or more of the image area of component j is in the marked area. =1, otherwise, =0, adjustment coefficient The initial value is set to 0.5, and the constant The value follows the rule that the larger the marked area, the smaller the value, and vice versa. >0;

[0063] The above formula is used to calculate the concern level of each marked area, which provides support for the subsequent evaluation of the circuit board reliability by the system in this embodiment;

[0064] An analysis module is configured to receive a virtual PCB three-dimensional model and a PCB outline image, collect a comparison analysis image group from the virtual PCB three-dimensional model and the PCB outline image, and analyze a reliability index of the PCB based on the comparison analysis image group;

[0065] The source of the comparative analysis image group collected by the analysis module is the marked area obtained by the marking unit and the regional contour image obtained by corresponding segmentation in the circuit board contour image based on the marked area;

[0066] Each comparison and analysis image group includes a marked area and an area outline image, and the marked area and area outline image included in the same comparison and analysis image group correspond to the same area on the circuit board;

[0067] The circuit board reliability indicators analyzed in the analysis module based on the comparison analysis image group are:

[0068] ;

[0069] Where: It is the reliability index of the circuit board; Annotated area image The gradient amplitude of is the region contour image; It is an adjustment parameter with a value range of 0.01~0.1; is the binary version of the labeled area image; Indicates finding the sum of squares of pixel differences within the brackets; is the total area of the labeled region image; Annotated area image Frequency domain representation of ; is the region contour image Frequency domain representation of ; is a very small positive number;

[0070] in, The larger it is, the more reliable the area contour image participating in the calculation corresponds to the area on the circuit board;

[0071] The circuit board reliability index is calculated through the above logical formula, providing necessary parameter support for the evaluation of circuit board reliability;

[0072] An evaluation module, configured to receive the analysis results of the reliability indicators of each circuit board from the analysis module, and evaluate the reliability of the circuit board based on the analysis results;

[0073] When evaluating the reliability of a circuit board based on the analysis results of the reliability indicators of each circuit board, the evaluation module introduces the attention level of the marked area corresponding to the analysis results of each reliability indicator into the evaluation process:

[0074] ;

[0075] Where: is the reliability performance value of the circuit board; The total amount of image groups was analyzed for control; is the control analysis image group obtained corresponding to the vth control analysis image group; for The attention level of the corresponding marked area; is the total amount of the marked area; is the attention level of the i-th marked area;

[0076] Through the above logic formula, the maximum performance value of the circuit board reliability is calculated to provide support for the operation of the judgment module;

[0077] A determination module is used to obtain the circuit board reliability evaluation result in the evaluation module, set a circuit board qualification determination threshold, compare the determination threshold with the circuit board reliability evaluation result, and determine that the circuit board is qualified when the circuit board reliability evaluation result is greater than or equal to the determination threshold;

[0078] A recording module is used to trigger operation when the judgment module determines that the circuit board is unqualified, and record relevant information about the defects of the unqualified circuit board;

[0079] During the recording module operation phase, the reliability index of each focus area corresponding to the unqualified circuit board is obtained, and each reliability index is further compared with the judgment threshold in the judgment module. The focus area where the reliability index is less than the judgment threshold is recorded as the record content, that is, the operation of recording the defect information of the unqualified circuit board;

[0080] The acquisition module is interactively connected to the upload module through a wireless network, the upload module is interactively connected to the selection unit and the marking unit through a wireless network, the upload module is interactively connected to the analysis module and the evaluation module through a wireless network, the evaluation module is interactively connected to the judgment module and the recording module through a wireless network, and the recording module is interactively connected to the marking unit through a wireless network.

[0081] In this embodiment, the acquisition module operates to acquire circuit board image data, converts the circuit board image data into a grayscale image, and extracts a circuit board outline image from the grayscale image. The upload module is then operated to upload circuit board structural parameters and circuit and component distribution information. A virtual circuit board three-dimensional model is constructed based on the circuit board structural parameters and circuit and component distribution information. A selection unit simultaneously selects a marked area on the side of the virtual circuit board three-dimensional model where the circuit and component distribution model is located. The marking unit obtains the marked area selected by the selection module in real time, marks and selects the marked area with an image frame, and configures the marking information. The analysis module then receives the virtual circuit board three-dimensional model and the circuit board outline image, acquires a comparison analysis image group from the virtual circuit board three-dimensional model and the circuit board outline image, and analyzes circuit board reliability indicators based on the comparison analysis image group. The evaluation module further receives analysis results for each circuit board reliability indicator from the analysis module and evaluates circuit board reliability based on each analysis result. Finally, the determination module obtains the circuit board reliability evaluation result from the evaluation module, sets a circuit board qualification determination threshold, compares the circuit board reliability evaluation result based on the determination threshold, and determines that the circuit board is qualified when the circuit board reliability evaluation result is greater than or equal to the determination threshold. The recording module is synchronously triggered to operate when the determination module determines that the circuit board is unqualified, and records information related to defects of the unqualified circuit board.

[0082] Through the operation of the system in the above embodiment, the appearance reliability of the circuit board is evaluated to ensure that the quality and performance of the finished circuit board meet the production requirements;

[0083] It should be noted that:

[0084] 1. Interaction process of modules in the system

[0085] Data flow logic: The PCB outline image output by the acquisition module must be aligned with the virtual PCB 3D model constructed by the upload module to ensure that the corresponding areas of the physical image and the virtual model accurately match. The reliability indicators generated by the analysis module must be weighted according to the attention level of the marked areas. The evaluation module comprehensively calculates the overall reliability performance value. The qualified threshold of the judgment module can be customized by the user and dynamically adjusted according to industry standards or product specifications.

[0086] Operation sequence: The system runs according to a fixed process: the acquisition module obtains image data → the upload module builds a three-dimensional model → the analysis module performs image comparison → the evaluation module calculates reliability → the judgment module outputs qualified results → the recording module triggers defect recording (if unqualified).

[0087] 2. Key Technical Details

[0088] Image acquisition and processing:

[0089] Collection requirements: The camera must meet a certain resolution and be photographed facing the component surface of the circuit board under uniform lighting to ensure grayscale image quality.

[0090] ROI mask definition: ROI (region of interest) masks can be manually selected by the user on the virtual 3D model, or automatically identified by algorithms for critical areas such as high-power components and dense traces.

[0091] Virtual 3D model construction:

[0092] Carrier model: This model must include physical parameters such as the PCB substrate material (e.g., FR-4), thickness, and number of layers (single, double, or multi-layer).

[0093] Component distribution model: The component package size and solder joint coordinate modeling accuracy (e.g., micron level) must be specified to ensure consistency with the actual layout.

[0094] Calculation of attention level of marked area:

[0095] Component importance weight: Users can preset weights based on component functions (e.g., main control chips have a high weight, passive components have a low weight).

[0096] Adjustment coefficient and constant: The adjustment coefficient can be adjusted according to the complexity of the circuit board (for example, a higher value is used for complex circuits), and the constant is inversely proportional to the area of the marked area.

[0097] Reliability index calculation:

[0098] Gradient amplitude algorithm: The Sobel operator or Canny edge detection algorithm can be used to calculate the image gradient and highlight the component edges and trace outlines.

[0099] Frequency domain analysis method: Convert the image to the frequency domain through Fourier transform, compare the frequency component differences between the marked area and the actual outline, and evaluate the rationality of the layout.

[0100] Example 2:

[0101] A method for evaluating the reliability of an electronic product comprises the following steps:

[0102] Step 1: Collect PCB image data, convert the PCB image data into a grayscale image, optimize the grayscale image, and extract the PCB outline image from the optimized grayscale image;

[0103] Step 2: Upload the circuit board structural parameters and circuit and component distribution information, and build a virtual circuit board 3D model based on the circuit board structural parameters and circuit and component distribution information;

[0104] Step 3: Acquire an image of the marked area on the surface of the virtual circuit board 3D model, compare the marked area segmentation in the circuit board contour image to obtain a regional contour image, and combine the marked area image with the regional contour image to obtain a comparative analysis image group;

[0105] Step 4: Analyze the reliability index of the circuit board based on each comparison analysis image group;

[0106] Step 5: Evaluate the reliability of the circuit board based on the circuit board reliability index analysis results of each control analysis image group;

[0107] Step 6: Set the judgment threshold and obtain the circuit board reliability evaluation result. Based on the comparison between the evaluation result and the judgment threshold, determine whether the circuit board is qualified;

[0108] Step 7: Identify and record defective areas on unqualified circuit boards.

[0109] In summary, the method and system in the above embodiment combine image acquisition with three-dimensional modeling to construct a comparative analysis system between the physical image and virtual model of the circuit board. The image clarity can be improved based on the grayscale image optimization algorithm, and the accurate digital mapping of the circuit board structure and component distribution can be achieved through multi-dimensional parameter modeling. The attention level evaluation model is introduced to dynamically calculate the regional attention priority based on parameters such as circuit density, component importance, and the number of vias. The reliability index is quantified by combining frequency domain analysis and pixel difference algorithm to make the evaluation results more in line with the actual defect risk. The regional attention level weight is integrated in the evaluation process to achieve graded reference judgment of key areas. At the same time, through the indicator backtracking mechanism of unqualified circuit boards, the source area of reliability parameters below the threshold is accurately located to provide data support for defect analysis, which is different from the traditional single image analysis technology.

[0110] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. An electronic product reliability assessment system, characterized in that: include: An acquisition module is used to acquire circuit board image data, convert the circuit board image data into a grayscale image, and extract a circuit board contour image from the grayscale image; An upload module is used to upload circuit board structural parameters and circuit and component distribution information, and to build a virtual circuit board three-dimensional model based on the circuit board structural parameters and circuit and component distribution information; The upload module is provided with a selection unit and a marking unit at the lower level. The selection unit is used to select a marking area on the side where the circuit and component distribution model are located on the virtual circuit board three-dimensional model. The marking unit is used to obtain the marking area selected in the selection module, mark and select the marking area with an image frame, and configure the marking information. During the annotation unit operation phase, image frame annotation is performed from the perspective of the side where the circuit and component distribution model is located on the three-dimensional model of the virtual circuit board. The size of the image frame is customized by the system user. After the image frame annotation is completed, the annotation information configured is the attention level of the annotation area. An analysis module is configured to receive a virtual PCB three-dimensional model and a PCB outline image, collect a comparison analysis image group from the virtual PCB three-dimensional model and the PCB outline image, and analyze a reliability index of the PCB based on the comparison analysis image group; An evaluation module, configured to receive the analysis results of the reliability indicators of each circuit board from the analysis module, and evaluate the reliability of the circuit board based on the analysis results; A determination module is used to obtain the circuit board reliability evaluation result in the evaluation module, set a circuit board qualification determination threshold, compare the determination threshold with the circuit board reliability evaluation result, and determine that the circuit board is qualified when the circuit board reliability evaluation result is greater than or equal to the determination threshold; A recording module is used to trigger operation when the judgment module determines that the circuit board is unqualified, and record relevant information about the defects of the unqualified circuit board; The attention level of each marked area is subject to: Where: G(S i ) is the attention level of the i-th marked area; ρ(S i ) is the circuit trace density in the i-th marked area; ρ max is the global circuit trace density of the source surface image of the annotated area; ω j is the importance weight of component j; I j is whether element j exists in the sub-image; λ is the adjustment coefficient; N via (S i ) is the number of vias in the i-th marked area; N junction (S i ) is the number of crossing points in the i-th marked area; N total is the total number of global vias and intersections in the source surface image of the labeled area; κ is a constant; Among them, G(S i ) value, the more attention the marked area deserves. Based on the above formula, the attention level of each marked area is calculated. The importance weight of each component is preset by the system end user. j The value is 0 or 1. If half or more of the image area of component j is in the marked area, then I j =1, otherwise, I j =0, the adjustment coefficient λ is initially set to 0.5, and the constant κ follows the principle that the larger the marked area, the smaller the value, and vice versa, the larger the value, and the constant κ>0.

2. The electronic product reliability evaluation system according to claim 1, characterized in that: The circuit board image data collected by the acquisition module is a plane image of one side of the circuit board and the distribution of components on the circuit board. After obtaining the circuit board contour image, the acquisition module stores the circuit board contour image; After the acquisition module acquires the circuit board image data and converts the circuit board image data into a grayscale image, it first optimizes the grayscale image and then performs an extraction operation; The optimization processing logic of the grayscale image is expressed as: Where: I opt is the grayscale image of the circuit board after optimization; I raw is the original circuit board grayscale image; μ is the global mean of the image; σ is the global standard deviation of the image; is the image gradient; M is the mask of the region of interest; k1, k2, k3, k4 are hyperparameters; Among them, the value ranges of each hyperparameter are k1∈[0.5,2], k1∈[0.5,1.5], k1∈[0.1,0.5], k1∈[0.1,0.3], and k1, k2, k3, and k4 are used to adjust the scaling ratio of the image grayscale value, control the overall offset of the image grayscale value, determine the degree of edge enhancement, and adjust the contrast of the ROI area, respectively.

3. The electronic product reliability evaluation system according to claim 1, wherein: When constructing the virtual PCB three-dimensional model, the upload module applies the PCB structural parameters to construct a carrier model of the circuits and components on the PCB, and further constructs a circuit and component distribution model on the PCB based on the circuit and component distribution information on the carrier model. The combination of the carrier module and the circuit and component distribution model on the PCB is recorded as the virtual PCB three-dimensional model. Among them, the solder joints of each circuit and component on the circuit board are all represented on the virtual circuit board three-dimensional model.

4. The electronic product reliability evaluation system according to claim 1, characterized in that: The source of the comparative analysis image group collected by the analysis module is the marked area obtained by the marking unit and the regional contour image obtained by corresponding segmentation in the circuit board contour image based on the marked area; Each of the comparison and analysis image groups includes a marked area and an area contour image, and the marked area and the area contour image included in the same comparison and analysis image group correspond to the same area on the circuit board; The circuit board reliability index analyzed by the analysis module based on the comparison analysis image group is: Where: Q is the circuit board reliability index; is the gradient amplitude of the labeled region image I; C is the region contour image; γ is the adjustment parameter, ranging from 0.01 to 0.1; I bin is the binary version of the labeled region image; SSD(·) represents the sum of squared differences of the pixels in the brackets; Area(I) is the total area of the labeled region image; FT(I) is the frequency domain representation of the labeled region image I; FT(C) is the frequency domain representation of the region contour image C; ε is a very small positive number; The larger Q is, the more reliable the area contour image participating in the calculation corresponds to the area on the circuit board.

5. The electronic product reliability evaluation system according to claim 1, characterized in that: When evaluating the reliability of a circuit board based on the reliability index analysis results of each circuit board, the evaluation module introduces the attention level of the marked area corresponding to each reliability index analysis result into the evaluation process: Where: F is the reliability performance value of the circuit board; u is the total amount of the control analysis image group; Q v is the control analysis image group obtained corresponding to the vth control analysis image group; Q v The attention level of the corresponding marked area; n is the total number of marked areas; G(S i ) is the attention level of the i-th marked area.

6. The electronic product reliability evaluation system according to claim 1, characterized in that: During the operation phase of the recording module, the reliability indicators of each focus area corresponding to the circuit board judged as unqualified are obtained, and each reliability indicator is further compared with the judgment threshold in the judgment module. The focus area from which the reliability indicator is less than the judgment threshold is recorded as the record content, that is, the operation of recording the information related to the defects of the unqualified circuit board is performed.

7. The electronic product reliability evaluation system according to claim 1, characterized in that: The acquisition module is interactively connected to the upload module via a wireless network, the upload module is interactively connected to the selection unit and the marking unit via a wireless network, the upload module is interactively connected to the analysis module and the evaluation module via a wireless network, the evaluation module is interactively connected to the determination module and the recording module via a wireless network, and the recording module is interactively connected to the marking unit via a wireless network.

8. A method for evaluating the reliability of an electronic product, wherein the method is an implementation method of the electronic product reliability evaluation system according to any one of claims 1 to 7, characterized in that: The following steps are involved: Step 1: Collect PCB image data, convert the PCB image data into a grayscale image, optimize the grayscale image, and extract the PCB outline image from the optimized grayscale image; Step 2: Upload the circuit board structural parameters and circuit and component distribution information, and build a virtual circuit board 3D model based on the circuit board structural parameters and circuit and component distribution information; Step 3: Acquire an image of the marked area on the surface of the virtual circuit board 3D model, compare the marked area segmentation in the circuit board contour image to obtain a regional contour image, and combine the marked area image with the regional contour image to obtain a comparative analysis image group; Step 4: Analyze the reliability index of the circuit board based on each comparison analysis image group; Step 5: Evaluate the reliability of the circuit board based on the circuit board reliability index analysis results of each control analysis image group; Step 6: Set the judgment threshold and obtain the circuit board reliability evaluation result. Based on the comparison between the evaluation result and the judgment threshold, determine whether the circuit board is qualified; Step 7: Identify and record defective areas on unqualified circuit boards.

Images