Defect detection method and system for integrated circuit packaging

Abstract

The invention discloses a defect detection method and system for integrated circuit packaging, and the method comprises the steps: obtaining a scanning image of an integrated circuit packaging sample, and analyzing the scanning image to determine a defect region of the integrated circuit packaging sample; multi-modal data of the defect area are collected and fused, and comprehensive defect information of the integrated circuit packaging sample is obtained; counting defect types of the integrated circuit packaging sample and performance influence parameters caused by various types of defects according to the comprehensive defect information; and screening out to-be-repaired defects and negligible defects according to the defect types and performance influence parameters caused by each type of defect, and generating a visual report. The comprehensive defect information is determined by positioning the defect area and detecting the multi-modal fusion data of the defect area, so that multiple types of defects of the integrated circuit packaging sample can be determined in an all-around manner, and the defect detection precision, reliability and detection efficiency are ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of circuit defect detection, and particularly to a method and system for defect detection of integrated circuit packaging. Background Art

[0002] At present, integrated circuit packaging is an important link in the integrated circuit manufacturing process, and its quality directly affects the performance and reliability of integrated circuits. With the development of integrated circuits towards high density, high performance, and miniaturization, the packaging process has become increasingly complex, and it has become more and more difficult to detect packaging defects. Existing defect detection methods are all based on single-site visual inspection by industrial cameras and defect detection and positioning based on single-modal data. Their detection capabilities for different types of defects are limited, and high-precision defect detection and filtering cannot be achieved, reducing the detection efficiency. Summary of the Invention

[0003] In view of the problems shown above, the present invention provides a method and system for defect detection of integrated circuit packaging to solve the problems in the background art, such as single-site visual inspection by industrial cameras and defect detection and positioning based on single-modal data, which have limited detection capabilities for different types of defects, cannot achieve high-precision defect detection and filtering, and reduce the detection efficiency.

[0004] A method for defect detection of integrated circuit packaging includes the following steps: Obtain a scanned image of an integrated circuit packaging sample, and analyze the scanned image to determine the defect area of the integrated circuit packaging sample; Collect multi-modal data of the defect area and perform fusion to obtain comprehensive defect information of the integrated circuit packaging sample; According to the comprehensive defect information, count the defect types of the integrated circuit packaging sample and the performance impact parameters caused by each type of defect; According to the defect types and the performance impact parameters caused by each type of defect, screen out the defects to be repaired and the negligible defects and generate a visualization report.

[0005] Preferably, the step of obtaining a scanned image of an integrated circuit packaging sample and analyzing the scanned image to determine the defect area of the integrated circuit packaging sample includes: Determine the structural parameters of the integrated circuit packaging sample, determine the layering information of the integrated circuit packaging sample according to the structural parameters, and determine the scanning method according to the layering information. The scanning methods include: internal scanning, surface scanning, and layered scanning; Scan the integrated circuit packaging sample by the scanning method to obtain a scanned image, and perform image enhancement preprocessing on the scanned image; Obtain the defect characteristics of different types of defects, determine the identification elements of different types of defects according to the defect characteristics, and determine the identification and positioning strategies of different types of defects according to the identification elements; Based on the identification and positioning strategies of different types of defects, analyze the scanned image through image processing software, and identify and locate the defect areas of the integrated circuit package sample according to the analysis results.

[0006] Preferably, collect and fuse the multimodal data of the defect area to obtain the comprehensive defect information of the integrated circuit package sample, including: Determine the area attributes of the defect area, where the area attributes include: internal area, surface area, and delamination area, and determine the data collection method according to the area attributes, where: Collect the surface topography of the surface defect area through a high-resolution optical microscope, and determine the two-dimensional optical surface image of the defect area according to the surface topography; Scan the internal defect area with an ultrasonic probe to obtain the ultrasonic defect reflection data of the internal defect area; Perform X-ray imaging on the delamination defect area to obtain the three-dimensional X-ray image of the delamination defect area; Align and register the optical two-dimensional image, ultrasonic defect reflection data, and X-ray three-dimensional image in space to generate multimodal fusion data; Extract the current defect characteristics from the multimodal fusion data through a feature extraction algorithm, and generate the comprehensive defect information of the integrated circuit package sample based on the current defect characteristics using data fusion technology.

[0007] Preferably, count the defect types of the integrated circuit package sample and the performance impact parameters caused by each type of defect according to the comprehensive defect information, including: Determine the defect type distribution of the integrated circuit package sample according to the comprehensive defect information, and determine the target defect type corresponding to each defect area according to the defect type distribution; Determine the defect depth information and defect size corresponding to each defect area based on the multimodal fusion data through a defect depth analysis model based on a deep learning algorithm; Determine the defect morphology and position parameters of each defect area according to the defect depth information and defect size, obtain the package function information of the integrated circuit package sample, and determine the performance parameters to be analyzed according to the package function information; Obtain experimental defect analysis data according to the performance parameters to be analyzed, and construct a defect-performance relationship model according to the experimental defect analysis data; Based on the defect morphology and position parameters of each defect area, determine the target defect type performance impact correlation parameters and performance impact descriptive statistical parameters of each defect area through the defect-performance relationship model.

[0008] Preferably, defective parts to be repaired and negligible defective parts are screened out according to the defective part type and the performance impact parameters caused by each type of defective part, and a visualization report is generated, including: Determine the quantitative impact evaluation parameters and qualitative impact evaluation parameters of each type of defective part according to the defective part type and the performance impact parameters caused by each type of defective part; Determine the screening criteria for defective parts to be repaired and negligible defective parts respectively, and determine the reasonable parameter range and reasonable parameter description of each performance index according to the screening criteria; Divide each type of defective part into defective parts to be repaired and negligible defective parts according to the quantitative impact evaluation parameters and qualitative impact evaluation parameters of each type of defective part and the reasonable parameter range and reasonable parameter description of the performance indexes of defective parts to be repaired and negligible defective parts respectively; Generate repair suggestions according to the division result and the target defective part form of the defective part to be repaired, and generate a visualization report according to the repair suggestions.

[0009] A defective part detection system for integrated circuit packaging, the system includes: An analysis module, configured to obtain a scanned image of an integrated circuit packaging sample, and analyze the scanned image to determine the defective part area of the integrated circuit packaging sample; An acquisition module, configured to collect multi-modal data of the defective part area and perform fusion to obtain comprehensive defective part information of the integrated circuit packaging sample; A statistics module, configured to count the defective part type of the integrated circuit packaging sample and the performance impact parameters caused by each type of defective part according to the comprehensive defective part information; A generation module, configured to screen out defective parts to be repaired and negligible defective parts according to the defective part type and the performance impact parameters caused by each type of defective part, and generate a visualization report.

[0010] Preferably, the analysis module includes: A first determination sub-module, configured to determine the structural parameters of the integrated circuit packaging sample, determine the layering information of the integrated circuit packaging sample according to the structural parameters, and determine the scanning method according to the layering information, where the scanning method includes: internal scanning, surface scanning, and layering scanning; A scanning sub-module, configured to scan the integrated circuit packaging sample by the scanning method, obtain a scanned image, and perform image enhancement preprocessing on the scanned image; A second determination sub-module, configured to obtain the defective part features of different types of defective parts, determine the identification elements of different types of defective parts according to the defective part features, and determine the identification and positioning strategy of different types of defective parts according to the identification elements; An analysis sub-module, configured to analyze the scanned image through image processing software based on the identification and positioning strategy of different types of defective parts, and identify and locate the defective part area of the integrated circuit packaging sample according to the analysis result.

[0011] Preferably, the obtaining module includes: A third determination sub-module, configured to determine the regional attributes of the defect area, where the regional attributes include: internal area, surface area, and delamination area, and determine the data acquisition method according to the regional attributes, where: Collect the surface topography of the surface defect area through a high-resolution optical microscope, and determine the optical surface two-dimensional image of the defect area according to the surface topography; Scan the internal defect area with an ultrasonic probe to obtain the ultrasonic defect reflection data of the internal defect area; Perform X-ray imaging on the delamination defect area to obtain the X-ray three-dimensional image of the delamination defect area; A first generation sub-module, configured to perform spatial alignment and registration on the optical two-dimensional image, the ultrasonic defect reflection data, and the X-ray three-dimensional image to generate multi-modal fusion data; A second generation sub-module, configured to extract the current defect features from the multi-modal fusion data through a feature extraction algorithm, and generate comprehensive defect information of the integrated circuit package sample based on the current defect features by using a data fusion technology.

[0012] Preferably, the statistics module includes: A fourth determination sub-module, configured to determine the defect type distribution of the integrated circuit package sample according to the comprehensive defect information, and determine the target defect type corresponding to each defect area according to the defect type distribution; A fifth determination sub-module, configured to determine the defect depth information and defect size corresponding to each defect area based on the multi-modal fusion data through a defect depth analysis model based on a deep learning algorithm; A sixth determination sub-module, configured to determine the defect morphology and position parameters of each defect area according to the defect depth information and defect size, obtain the package function information of the integrated circuit package sample, and determine the performance parameters to be analyzed according to the package function information; A construction sub-module, configured to obtain experimental defect analysis data according to the performance parameters to be analyzed, and construct a defect-performance relationship model according to the experimental defect analysis data; A seventh determination sub-module, configured to determine the target defect type performance impact correlation parameter and performance impact descriptive statistical parameter of each defect area based on the defect morphology and position parameters of each defect area through the defect-performance relationship model.

[0013] Preferably, the generation module includes: An eighth determination sub-module, configured to determine the quantitative impact evaluation parameter and qualitative impact evaluation parameter of each type of defect according to the defect type and the performance impact parameters caused by each type of defect; A ninth determination sub-module, configured to determine the screening criteria for the defects to be repaired and the negligible defects respectively, and determine the reasonable parameter intervals and reasonable parameter descriptions of each performance index according to the screening criteria; A division sub-module, configured to divide each type of defect into defects to be repaired and negligible defects according to the quantitative impact evaluation parameters and qualitative impact evaluation parameters of each type of defect, and the reasonable parameter intervals and reasonable parameter descriptions of the performance indexes of the defects to be repaired and the negligible defects respectively; A third generation sub-module, configured to generate a repair suggestion according to the division result and the target defect form of the defect to be repaired, and generate a visualization report according to the repair suggestion.

[0014] Other features and advantages of the present invention will be described in the subsequent description, and, in part, will be obvious from the description, or will be understood by implementing the present invention. The objectives and other advantages of the present invention can be achieved and obtained by the structures specifically pointed out in the written description and the drawings.

[0015] The technical solution of the present invention will be further described in detail below through the drawings and embodiments. Description of the Drawings

[0016] The drawings are used to provide a further understanding of the present invention, and constitute a part of the description. They are used to explain the present invention together with the embodiments of the present invention, and do not constitute a limitation to the present invention.

[0017] Figure 1 It is a working flow chart of a defect detection method for an integrated circuit package provided by the present invention; Figure 2 It is another working flow chart of a defect detection method for an integrated circuit package provided by the present invention; Figure 3 It is a structural schematic diagram of a defect detection system for an integrated circuit package provided by the present invention; Figure 4 It is a structural schematic diagram of an analysis module in a defect detection system for an integrated circuit package provided by the present invention. Detailed Embodiments

[0018] Here, the exemplary embodiments will be described in detail, and the examples are shown in the drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. On the contrary, they are only examples of devices and methods consistent with some aspects of the present disclosure as detailed in the appended claims.

[0019] At present, integrated circuit packaging is an important link in the integrated circuit manufacturing process, and its quality directly affects the performance and reliability of integrated circuits. With the development of integrated circuits towards high density, high performance, and miniaturization, the packaging process has become increasingly complex, and packaging defects have become increasingly difficult to detect. Existing defect detection methods are all based on single-site visual inspection using industrial cameras and defect detection and localization are carried out based on single-modal data. Their detection capabilities for different types of defects are limited, and high-precision defect detection and filtering cannot be achieved, reducing the detection efficiency. To solve the above problems, this embodiment discloses a defect detection method for integrated circuit packaging.

[0020] A defect detection method for integrated circuit packaging, as Figure 1 shown, includes the following steps: Step S101: Obtain a scanned image of an integrated circuit packaging sample, and analyze the scanned image to determine the defect area of the integrated circuit packaging sample; Step S102: Collect multi-modal data of the defect area and perform fusion to obtain the comprehensive defect information of the integrated circuit packaging sample; Step S103: According to the comprehensive defect information, count the defect types of the integrated circuit packaging sample and the performance impact parameters caused by each type of defect; Step S104: Screen out the defects to be repaired and the negligible defects according to the defect types and the performance impact parameters caused by each type of defect, and generate a visualization report. In this embodiment, the defect types include: packaging defects, assembly defects, and material defects. Among them, packaging defects include: poor wire bonding, chip cracking, voids in packaging materials, etc.; assembly defects include solder joint voids, solder ball bridging, component misalignment, etc.; material defects include substrate cracks, solder impurities, delamination of plastic encapsulation materials, etc.

[0021] The working principle of the above technical solution is: obtain a scanned image of an integrated circuit packaging sample, analyze the scanned image to determine the defect area of the integrated circuit packaging sample; collect multi-modal data of the defect area and perform fusion to obtain the comprehensive defect information of the integrated circuit packaging sample; according to the comprehensive defect information, count the defect types of the integrated circuit packaging sample and the performance impact parameters caused by each type of defect; screen out the defects to be repaired and the negligible defects according to the defect types and the performance impact parameters caused by each type of defect, and generate a visualization report.

[0022] The beneficial effects of the above technical solution are as follows: By locating the defect area and detecting the multi-modal fusion data of the defect area to determine the comprehensive defect information, various types of defects in the integrated circuit package sample can be determined in all aspects, ensuring the accuracy, reliability, and detection efficiency of defect detection. Further, by evaluating the performance impact of each type of detected defect and then screening out negligible defects, defects with little performance impact can be filtered out, improving the defect repair efficiency, and solving the problem mentioned in the prior art that single-site visual inspection is performed by an industrial camera and defect detection and positioning are based on single-modal data, which has limited detection capabilities for different types of defects, cannot achieve high-precision defect detection and filtering, and reduces the detection efficiency.

[0023] In one embodiment, as Figure 2 shown, the steps of obtaining a scanned image of the integrated circuit package sample and analyzing the scanned image to determine the defect area of the integrated circuit package sample include: Step S201: Determine the structural parameters of the integrated circuit package sample, determine the layer information of the integrated circuit package sample according to the structural parameters, and determine the scanning method according to the layer information. The scanning methods include: internal scanning, surface scanning, and layer-by-layer scanning; Step S202: Scan the integrated circuit package sample through the scanning method, obtain a scanned image, and perform image enhancement preprocessing on the scanned image; Step S203: Obtain the defect characteristics of different types of defects, determine the identification elements of different types of defects according to the defect characteristics, and determine the identification and positioning strategies of different types of defects according to the identification elements; Step S204: Analyze the scanned image through image processing software based on the identification and positioning strategies of different types of defects, and identify and locate the defect area of the integrated circuit package sample according to the analysis results.

[0024] The beneficial effects of the above technical solution are as follows: By collecting different scanning methods to perform systematic scanning of the integrated circuit package sample, defect scanning can be achieved in all aspects, improving the scanning efficiency and comprehensiveness. Further, by positioning the defect area based on the identification and positioning strategies of different types of defects, accurate defect area positioning can be quickly performed through reference features, improving the positioning accuracy and positioning efficiency.

[0025] In this embodiment, obtaining the defect characteristics of different types of defects, determining the identification elements of different types of defects according to the defect characteristics, and determining the identification and positioning strategies of different types of defects according to the identification elements includes: Obtain the defect image dominant features and defect text description features of different types of defects, and determine the defect feature point distribution map according to the defect image dominant features and defect text description features; Determine the affine anomaly frequency components according to the defective feature point distribution diagram through the affine transformation matrix, compare the affine anomaly frequency components with the standard frequency components, and obtain the comparison result; Determine the recognition elements of different types of defects according to the comparison result and the intensifying factors of the anomaly frequency components, and obtain the superficial detection expression attributes and the depth detection expression attributes of the recognition elements; Respectively, for the superficial detection expression attributes and the depth detection expression attributes, their respective characteristic change arrays are obtained, and based on the characteristic change arrays, the low-frequency scanning parameters and the high-matching pair parameters for the recognition elements are determined; Determine the scanning form according to the low-frequency scanning parameters, and based on the scanning form, determine the low-frequency parameter recognition and positioning strategy for the recognition elements; Obtain the parameter change rules of the high-frequency matching parameters, determine the pairing mechanism according to the parameter change rules, and based on the pairing mechanism, determine the high-frequency parameter recognition and positioning strategy for the recognition elements; Integrate the low-frequency parameter recognition and positioning strategy and the high-frequency parameter recognition and positioning strategy for the recognition elements to generate the recognition and positioning strategy for different types of defects.

[0026] The beneficial effects of the above technical solutions are as follows: By determining the parameter recognition and positioning strategies for the superficial and deep levels of the recognition elements of different types of defects to comprehensively perform defect recognition and positioning, the recognition accuracy and recognition reliability for various types of defects can be fully guaranteed, and the accuracy and reliability of defect positioning and recognition are improved.

[0027] In one embodiment, the multi-modal data of the defective area is collected and fused to obtain the comprehensive defect information of the integrated circuit package sample, including: Determine the area attributes of the defective area, where the area attributes include: internal area, surface area, and delamination area, and determine the data collection method according to the area attributes, where: Collect the surface topography of the surface defective area through a high-resolution optical microscope, and determine the optical surface two-dimensional image of the defective area according to the surface topography; Scan the internal defective area with an ultrasonic probe to obtain the ultrasonic defect reflection data of the internal defective area; Perform X-ray imaging on the delamination defective area to obtain the X-ray three-dimensional image of the delamination defective area; Perform spatial alignment and registration on the optical two-dimensional image, the ultrasonic defect reflection data, and the X-ray three-dimensional image to generate multi-modal fusion data; Extract the current defect features from the multi-modal fusion data through a feature extraction algorithm, and use the data fusion technology to generate the comprehensive defect information of the integrated circuit package sample based on the current defect features.

[0028] The beneficial effects of the above technical solution are as follows: By performing multi-modal data collection and fusion, the structural parameters of each structure of the circuit packaging sample can be comprehensively obtained, and then defect information can be detected, improving the accuracy and efficiency of defect detection.

[0029] In one embodiment, according to the comprehensive defect information, the defect types of the integrated circuit packaging sample and the performance impact parameters caused by each type of defect are statistically analyzed, including: Determine the defect type distribution of the integrated circuit packaging sample according to the comprehensive defect information, and determine the target defect type corresponding to each defect area according to the defect type distribution; Based on the defect depth analysis model based on the deep learning algorithm, determine the defect depth information and defect size corresponding to each defect area based on the multi-modal fusion data; Determine the defect morphology and position parameters of each defect area according to the defect depth information and defect size, obtain the packaging function information of the integrated circuit packaging sample, and determine the performance parameters to be analyzed according to the packaging function information; Obtain experimental defect analysis data according to the performance parameters to be analyzed, and construct a defect-performance relationship model according to the experimental defect analysis data; Based on the defect-performance relationship model, determine the performance impact correlation parameters and performance impact descriptive statistical parameters of the target defect type for each defect area based on the defect morphology and position parameters of each defect area.

[0030] The beneficial effects of the above technical solution are as follows: By constructing a defect-performance relationship model, the performance impact evaluation process can be quickly carried out according to the defect forms of each type of defect, improving the evaluation efficiency and accuracy. Further, by determining the performance impact correlation parameters and performance impact descriptive statistical parameters of the target defect type for each defect area, the performance impact description parameters and correlation parameter impact description parameters of each type of defect on the operating performance of the integrated circuit packaging sample can be accurately determined, comprehensively determining the impact of performance, laying a reference condition for subsequent defect screening and repair recommendation generation, and improving the practicality.

[0031] In one embodiment, according to the defect type and the performance impact parameters caused by each type of defect, the defects to be repaired and the negligible defects are screened out and a visualization report is generated, including: Determine the quantitative impact evaluation parameters and qualitative impact evaluation parameters of each type of defect according to the defect type and the performance impact parameters caused by each type of defect; Determine the screening criteria for the defects to be repaired and the negligible defects respectively, and determine the reasonable parameter interval and reasonable parameter description of each performance index according to the screening criteria; Classify various types of defects into defects to be repaired and negligible defects according to the quantitative impact evaluation parameters and qualitative impact evaluation parameters of each type of defect, as well as the reasonable parameter intervals and reasonable parameter descriptions of the performance indicators of the defects to be repaired and negligible defects respectively; Generate repair suggestions according to the classification results and the target defect morphology of the defects to be repaired, and generate a visualization report according to the repair suggestions.

[0032] The beneficial effects of the above technical solution are: it can visually and quickly classify the detected defects in each defect area and then quickly generate repair suggestions, improving work efficiency.

[0033] In one embodiment, this embodiment also discloses a defect detection system for integrated circuit packaging, as Figure 3 shown, the system includes: An analysis module 301, configured to obtain a scanned image of an integrated circuit packaging sample, and analyze the scanned image to determine the defect area of the integrated circuit packaging sample; An acquisition module 302, configured to collect multi-modal data of the defect area and perform fusion to obtain comprehensive defect information of the integrated circuit packaging sample; A statistics module 303, configured to count the defect types of the integrated circuit packaging sample and the performance impact parameters caused by each type of defect according to the comprehensive defect information; A generation module 304, configured to screen out defects to be repaired and negligible defects according to the defect types and the performance impact parameters caused by each type of defect, and generate a visualization report.

[0034] The working principle and beneficial effects of the above technical solution have been described in the method embodiment, and will not be elaborated here.

[0035] In one embodiment, as Figure 4 shown, the analysis module 301 includes: A first determination sub-module 3011, configured to determine the structural parameters of the integrated circuit packaging sample, determine the layering information of the integrated circuit packaging sample according to the structural parameters, and determine the scanning method according to the layering information, where the scanning method includes: internal scanning, surface scanning, and layer-by-layer scanning; A scanning sub-module 3012, configured to scan the integrated circuit packaging sample by the scanning method, obtain a scanned image, and perform image enhancement preprocessing on the scanned image; A second determination sub-module 3013, configured to obtain the defect characteristics of different types of defects, determine the recognition elements of different types of defects according to the defect characteristics, and determine the recognition and positioning strategies of different types of defects according to the recognition elements; The analysis sub-module 3014 is used to analyze the scanned image through image processing software based on the identification and localization strategies of different types of defects, and identify and locate the defect areas of the integrated circuit package sample according to the analysis results.

[0036] In one embodiment, the acquisition module includes: The third determination sub-module is used to determine the regional attributes of the defect area. The regional attributes include: internal area, surface area, and delamination area, and determine the data acquisition method according to the regional attributes, where: Collect the surface topography of the surface defect area through a high-resolution optical microscope, and determine the optical surface two-dimensional image of the defect area according to the surface topography; Scan the internal defect area with an ultrasonic probe to obtain the ultrasonic defect reflection data of the internal defect area; Perform X-ray imaging on the delamination defect area to obtain the X-ray three-dimensional image of the delamination defect area; The first generation sub-module is used to spatially align and register the optical two-dimensional image, the ultrasonic defect reflection data, and the X-ray three-dimensional image to generate multi-modal fusion data; The second generation sub-module is used to extract the current defect features from the multi-modal fusion data through a feature extraction algorithm, and generate the comprehensive defect information of the integrated circuit package sample based on the current defect features using data fusion technology.

[0037] In one embodiment, the statistics module includes: The fourth determination sub-module is used to determine the defect type distribution of the integrated circuit package sample according to the comprehensive defect information, and determine the target defect type corresponding to each defect area according to the defect type distribution; The fifth determination sub-module is used to determine the defect depth information and defect size corresponding to each defect area based on the multi-modal fusion data through a defect depth analysis model based on a deep learning algorithm; The sixth determination sub-module is used to determine the defect morphology and position parameters of each defect area according to the defect depth information and defect size, obtain the package function information of the integrated circuit package sample, and determine the performance parameters to be analyzed according to the package function information; The construction sub-module is used to obtain experimental defect analysis data according to the performance parameters to be analyzed, and construct a defect-performance relationship model according to the experimental defect analysis data; The seventh determination sub-module is used to determine the target defect type performance impact correlation parameters and performance impact descriptive statistical parameters of each defect area based on the defect morphology and position parameters of each defect area through the defect-performance relationship model.

[0038] In one embodiment, the generation module includes: The eighth determination sub-module is configured to determine the quantitative impact evaluation parameter and the qualitative impact evaluation parameter of each type of defect according to the defect type and the performance impact parameter caused by each type of defect; The ninth determination sub-module is configured to determine the screening criteria for the defects to be repaired and the negligible defects respectively, and determine the reasonable parameter interval and reasonable parameter description of each performance index according to the screening criteria; The division sub-module is configured to divide each type of defect into defects to be repaired and negligible defects according to the quantitative impact evaluation parameter and the qualitative impact evaluation parameter of each type of defect, and the reasonable parameter interval and reasonable parameter description of the performance indexes of the defects to be repaired and the negligible defects respectively; The third generation sub-module is configured to generate a repair suggestion according to the division result and the target defect form of the defect to be repaired, and generate a visualization report according to the repair suggestion.

[0039] Those skilled in the art should understand that the first and second in the present invention refer to different application stages only.

[0040] After considering the specification and practicing the disclosure herein, those skilled in the art will readily conceive of other embodiments of the present disclosure. This application is intended to cover any variations, uses, or adaptations of the present disclosure that follow the general principles of the present disclosure and include known common knowledge or conventional technical means in the technical field not disclosed in the present disclosure. The specification and examples are only illustrative, and the true scope and spirit of the present disclosure are pointed out by the following claims.

[0041] It should be understood that the present disclosure is not limited to the exact structures described above and shown in the drawings, and various modifications and changes can be made without departing from its scope. The scope of the present disclosure is only limited by the appended claims.

Claims

1. A defect detection method for integrated circuit packaging, characterized in that: The following steps are involved: Acquire a scanned image of the integrated circuit package sample, and analyze the scanned image to determine a defective area of ​​the integrated circuit package sample; Collect multimodal data of defective areas and fuse them to obtain comprehensive defect information of integrated circuit packaging samples; Count the defect types of integrated circuit package samples and the performance impact parameters caused by each type of defect based on the comprehensive defect information; According to the defect type and the performance impact parameters caused by each type of defect, defects to be repaired and defects that can be ignored are screened out and a visual report is generated.

2. The defect detection method for integrated circuit packaging according to claim 1, characterized in that: The step of acquiring a scanned image of the integrated circuit package sample and analyzing the scanned image to determine a defective area of ​​the integrated circuit package sample comprises: Determine structural parameters of the integrated circuit package sample, determine layer information of the integrated circuit package sample according to the structural parameters, and determine a scanning method according to the layer information, wherein the scanning method includes: internal scanning, surface scanning and layer scanning; Scanning the integrated circuit package sample by scanning, acquiring the scanned image, and performing image enhancement preprocessing on the scanned image; Obtain defect characteristics of different types of defects, determine identification factors of different types of defects based on the defect characteristics, and determine identification and positioning strategies of different types of defects based on the identification factors; The scanned image is analyzed by image processing software based on the identification and positioning strategies of different types of defects, and the defective areas of the integrated circuit packaging samples are identified and located according to the analysis results.

3. The defect detection method for integrated circuit packaging according to claim 1, characterized in that: The multi-modal data of the defective area is collected and fused to obtain comprehensive defect information of the integrated circuit packaging sample, including: Determine the regional attributes of the defect area, the regional attributes include: internal area, surface area and layered area, and determine the data collection method according to the regional attributes, wherein: The surface morphology of the surface defect area is collected by a high-resolution optical microscope, and a two-dimensional optical surface image of the defect area is determined according to the surface morphology; Scan the internal defect area with an ultrasonic probe to obtain ultrasonic defect reflection data of the internal defect area; Performing X-ray imaging on the delamination defect area to obtain an X-ray three-dimensional image of the delamination defect area; Spatially align and register the optical two-dimensional image, ultrasonic defect reflection data, and X-ray three-dimensional image to generate multi-modal fusion data; The current defect features are extracted from the multimodal fusion data through the feature extraction algorithm, and the comprehensive defect information of the integrated circuit packaging sample is generated based on the current defect features using data fusion technology.

4. The defect detection method for integrated circuit packaging according to claim 3, characterized in that: According to the comprehensive defect information, the defect types of integrated circuit packaging samples and the performance impact parameters caused by each type of defect are counted, including: Determine the defect type distribution of the integrated circuit package sample according to the comprehensive defect information, and determine the target defect type corresponding to each defect area according to the defect type distribution; Determine the defect depth information and defect size corresponding to each defect area based on multimodal fusion data through a defect depth analysis model based on a deep learning algorithm; Determine the defect morphology and position parameters of each defect area according to the defect depth information and defect size, obtain the packaging function information of the integrated circuit packaging sample, and determine the performance parameters to be analyzed according to the packaging function information; Acquire experimental defect analysis data according to the performance parameters to be analyzed, and construct a defect-performance relationship model according to the experimental defect analysis data; The defect-performance relationship model is used to determine the target defect type performance impact correlation parameters and performance impact descriptive statistical parameters of each defect area based on the defect morphology and location parameters of each defect area.

5. The defect detection method for integrated circuit packaging according to claim 4, characterized in that: According to the defect type and the performance impact parameters caused by each type of defect, defects to be repaired and defects that can be ignored are screened out and a visual report is generated, including: Determine the quantitative impact assessment parameters and qualitative impact assessment parameters of each type of defect according to the defect type and the performance impact parameters caused by each type of defect; Determine the screening criteria for defects to be repaired and defects that can be ignored, and determine the reasonable parameter range and reasonable parameter description of each performance indicator based on the screening criteria; According to the quantitative impact assessment parameters and qualitative impact assessment parameters of each type of defect and the reasonable parameter range and reasonable parameter description of the performance indicators of the defects to be repaired and the defects to be ignored, each type of defect is divided into defects to be repaired and defects to be ignored; Generate repair suggestions based on the segmentation results and the target defect morphology of the defect to be repaired, and generate a visual report based on the repair suggestions.

6. A defect detection system for integrated circuit packaging, characterized in that: The system includes: An analysis module, used to obtain a scanned image of the integrated circuit package sample, and analyze the scanned image to determine a defective area of ​​the integrated circuit package sample; An acquisition module is used to collect and fuse multimodal data of defective areas to obtain comprehensive defect information of integrated circuit packaging samples; A statistical module, used to count the defect types of integrated circuit packaging samples and the performance impact parameters caused by each type of defect based on the comprehensive defect information; The generation module is used to filter out defects to be repaired and defects that can be ignored according to the defect type and the performance impact parameters caused by each type of defect, and generate a visual report.

7. The defect detection system for integrated circuit packaging according to claim 6, characterized in that: The analysis module comprises: A first determination submodule is used to determine the structural parameters of the integrated circuit package sample, determine the layer information of the integrated circuit package sample according to the structural parameters, and determine the scanning mode according to the layer information, wherein the scanning mode includes: internal scanning, surface scanning and layer scanning; The scanning submodule is used to scan the integrated circuit package sample by scanning, obtain the scanned image, and perform image enhancement preprocessing on the scanned image; The second determination submodule is used to obtain defect characteristics of different types of defects, determine identification elements of different types of defects according to the defect characteristics, and determine identification and positioning strategies of different types of defects according to the identification elements; The analysis submodule is used to analyze the scanned image through image processing software based on the identification and positioning strategies of different types of defects, and identify and locate the defective area of ​​the integrated circuit packaging sample according to the analysis results.

8. The integrated circuit package defect detection system according to claim 6, characterized in that: The acquisition module comprises: The third determination submodule is used to determine the regional attributes of the defect area, the regional attributes include: internal area, surface area and layered area, and determine the data collection method according to the regional attributes, wherein: The surface morphology of the surface defect area is collected by a high-resolution optical microscope, and a two-dimensional optical surface image of the defect area is determined according to the surface morphology; Scan the internal defect area with an ultrasonic probe to obtain ultrasonic defect reflection data of the internal defect area; Performing X-ray imaging on the delamination defect area to obtain an X-ray three-dimensional image of the delamination defect area; The first generation submodule is used to spatially align and register the optical two-dimensional image, ultrasonic defect reflection data, and X-ray three-dimensional image to generate multimodal fusion data; The second generation submodule is used to extract the current defect features from the multimodal fusion data through a feature extraction algorithm, and generate comprehensive defect information of the integrated circuit packaging sample based on the current defect features using data fusion technology.

9. The defect detection system for integrated circuit packaging according to claim 8, characterized in that: Statistics module, including: A fourth determination submodule is used to determine the defect type distribution of the integrated circuit package sample according to the comprehensive defect information, and determine the target defect type corresponding to each defect area according to the defect type distribution; A fifth determination submodule is used to determine the defect depth information and defect size corresponding to each defect area based on the multimodal fusion data through a defect depth analysis model based on a deep learning algorithm; The sixth determination submodule is used to determine the defect morphology and position parameters of each defect area according to the defect depth information and the defect size, obtain the packaging function information of the integrated circuit packaging sample, and determine the performance parameter to be analyzed according to the packaging function information; A construction submodule is used to obtain experimental defect analysis data according to the performance parameters to be analyzed, and to construct a defect-performance relationship model according to the experimental defect analysis data; The seventh determination submodule is used to determine the target defect type performance impact correlation parameter and the performance impact descriptive statistical parameter of each defect area based on the defect morphology and position parameters of each defect area through the defect-performance relationship model.

10. The integrated circuit package defect detection system according to claim 9, characterized in that: The generating module comprises: An eighth determination submodule, configured to determine quantitative impact assessment parameters and qualitative impact assessment parameters of each type of defect according to the defect type and the performance impact parameters caused by each type of defect; The ninth determination submodule is used to determine the screening criteria of defects to be repaired and defects that can be ignored, and determine the reasonable parameter range and reasonable parameter description of each performance indicator according to the screening criteria; A classification submodule is used to classify each type of defect into defects to be repaired and negligible defects according to the quantitative impact assessment parameters and qualitative impact assessment parameters of each type of defect and the reasonable parameter range and reasonable parameter description of the performance indicators of the defects to be repaired and the defects to be ignored; The third generation submodule is used to generate repair suggestions according to the division results and the target defect morphology of the defect to be repaired, and generate a visualization report according to the repair suggestions.

Images