Connector detection method, device, computer equipment and storage medium

Abstract

This invention relates to a connector inspection method, apparatus, computer device, and storage medium. By combining high-frequency ultrasonic testing and optical coherence tomography (OCT) technology, it provides a high-precision, comprehensive connector inspection method. Specific contributions are as follows: The combination of high-frequency ultrasonic testing and OCT enables the inspection method to not only obtain surface information about the connector but also penetrate deep into the interior for comprehensive detection of internal defects. The high-resolution three-dimensional images provided by OCT accurately identify tiny defects, improving inspection accuracy and reliability. Image processing and analysis algorithms can extract detailed defect parameters, providing more specific and useful inspection information. Comprehensive analysis of defect characteristics allows for a systematic assessment of connector quality, generating detailed inspection reports that support quality control and improvement.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a connector detection method, device, computer equipment and storage medium. Background Art

[0002] In electronics manufacturing, connector quality is crucial to the performance and stability of the entire electronic device. Connector functionality depends primarily on the interface parameters between its contact terminals and insulators, which require precise inspection to ensure high quality and reliability. Traditional inspection methods rely primarily on visual inspection and manual measurement, or use simple two-dimensional image processing techniques to identify and analyze key connector features. While these methods provide basic means of measuring electrical performance and physical dimensions, the increasing miniaturization and sophistication of electronic devices places increasing demands on precision and comprehensiveness.

[0003] Conventional methods typically use two-dimensional image analysis to inspect connector interface parameters. This involves capturing a two-dimensional image of the connector using a camera and preprocessing the captured image, including noise reduction and enhancement. Next, algorithms such as edge detection and morphological processing are used to extract key features of the contact terminals and insulator, such as size, position, and shape. Finally, the extracted features are analyzed to determine whether the connector's electrical performance and physical dimensions meet standards.

[0004] However, the limitation of 2D image analysis is that it only provides two-dimensional information and cannot fully reflect the connector's internal defects and microscopic topography. Therefore, defects such as the micromorphology of contact terminals and hidden structural defects within the insulator may not be detected. This results in insufficient detection accuracy and reliability, making it difficult to meet the requirements of modern high-precision and complex connector quality inspection. Summary of the Invention

[0005] The main purpose of the present invention is to provide a connector detection method, device, computer equipment and storage medium to solve the problem in the prior art that the detection of connector interface parameters usually adopts a two-dimensional image analysis method, which cannot fully reflect the internal defects and micromorphology of the connector, resulting in insufficient detection accuracy and reliability, and it is difficult to meet the needs of modern high-precision and complex connector quality inspection.

[0006] To achieve the above-mentioned objectives, the present invention provides a method for detecting a connector, which includes the following steps: using a high-frequency ultrasonic detection device to perform an omnidirectional scan of the connector to obtain an ultrasonic reflection image of the connector; using an optical coherence tomography device to perform high-resolution surface and internal imaging of the connector to obtain an OCT three-dimensional image of the connector; performing defect identification and feature extraction based on the ultrasonic reflection image and the OCT three-dimensional image of the connector to generate a defect feature list and detailed parameter information of each defect feature; and performing detection and evaluation of the connector based on the defect feature list and the detailed parameter information of each defect feature to obtain a detection and evaluation data report of the connector.

[0007] Furthermore, a high-frequency ultrasonic detection device is used to perform an all-round scan of the connector to obtain an ultrasonic reflection image of the connector, including: scanning the connector with a high-frequency ultrasonic probe to collect initial ultrasonic reflection data from different angles and depths; performing denoising and signal enhancement processing on the initial ultrasonic reflection data to eliminate background noise in the initial ultrasonic reflection data and enhance the signal strength of the initial ultrasonic reflection data to obtain preprocessed ultrasonic reflection data; performing a Hilbert transform on the preprocessed ultrasonic reflection data to obtain the imaginary part and the real part of the complex envelope signal, and calculating the envelope signal based on the imaginary part and the real part of the complex envelope signal; performing normalization processing on the envelope signal and converting the normalized envelope signal into an image format so that each value in the envelope signal is mapped to the grayscale or color value of the image to obtain an ultrasonic reflection image.

[0008] Furthermore, an optical coherence tomography device is used to perform high-resolution surface and internal imaging of the connector to obtain an OCT three-dimensional image of the connector, including: scanning the connector using an optical coherence tomography scanner to obtain microstructural information on the surface and interior of the connector to obtain an original OCT data set; preprocessing the original OCT data set to sequentially eliminate background noise of the original OCT data set, correct depth signal attenuation of the original OCT data set, and correct phase error of the original OCT data set to obtain initial OCT data; using trilinear interpolation to calculate voxel values ​​of the initial OCT data, and based on the voxel values ​​of the initial OCT data, performing voxel reconstruction processing on the initial OCT data in combination with ray casting and maximum intensity projection to convert the two-dimensional data into three-dimensional data, and finally obtaining a reconstructed OCT three-dimensional image.

[0009] Furthermore, defect recognition and feature extraction are performed based on the ultrasonic reflection image of the connector to generate a list of defect features and detailed parameter information of each defect feature, including: removing high-frequency noise and low-frequency noise in the ultrasonic reflection image through a high-pass filter and a low-pass filter, and using a contrast enhancement algorithm to improve the clarity of the ultrasonic reflection image, thereby obtaining an initial ultrasonic reflection image in which the internal structure of the connector is more obvious; using a preset segmentation method based on region growing to segment the initial ultrasonic reflection image into multiple regions, wherein each region corresponds to a different component of the connector, and the components include at least: contact terminals and insulators; The method further comprises the following steps: performing feature extraction on the image area after the initial ultrasonic reflection image to obtain the internal and external morphological features and internal and external texture features of each component of the connector, and adopting a machine learning algorithm based on a support vector machine to compare the extracted features with a preset connector defect feature library to identify the defect features represented by the initial ultrasonic reflection image and detailed parameter information of each defect feature; generating a defect feature list based on the identified defect features, storing the defect feature list in a structured data format, and recording the detailed parameter information of each defect feature in the defect feature list, wherein the detailed parameter information includes at least type data, position data, size data and severity information of the defect feature.

[0010] Furthermore, defect recognition and feature extraction are performed based on the OCT three-dimensional image of the connector to generate the defect feature list and detailed parameter information of each defect feature, including: removing high-frequency noise and low-frequency noise in the OCT three-dimensional image by an adaptive stop filtering method, and using a histogram equalization method to improve the image contrast of the OCT three-dimensional image to obtain an initial OCT three-dimensional image representing the internal and external structure of the connector; using a preset segmentation method based on region growing to segment the initial OCT three-dimensional image into multiple regions, wherein each region corresponds to a different component of the connector, and the components include at least: contact terminals and insulators; the segmented image The method comprises the following steps: performing feature extraction on the region to obtain the internal and external morphological features and internal and external texture features of each component of the connector, and using a machine learning algorithm based on a support vector machine to compare the extracted features with a preset connector defect feature library to identify the defect features of the connector represented by the initial OCT three-dimensional image, as well as detailed parameter information of each defect feature; generating a defect feature list based on the identified defect features, storing the defect feature list in a structured data format, and recording the detailed parameter information of each defect feature in the defect feature list, wherein the detailed parameter information includes at least type data, position data, size data and severity information of the defect feature.

[0011] Furthermore, the connector is inspected and evaluated based on the defect feature list and the detailed parameter information of each defect feature to obtain an inspection and evaluation data report of the connector, including: classifying the defect features of the connector according to the defect type field in the defect feature list, and extracting the position data, size data and severity information of each defect feature in each type of defect feature in turn; calculating the area ratio value and defect depth value of each defect feature, as well as the overall defect distribution value and overall defect density value of the connector based on the position data and size data of each defect feature; determining the weight distribution value corresponding to each defect feature based on the severity information of each defect feature, and performing weighted calculation on the area ratio value and defect depth value of each defect feature based on the weight distribution value corresponding to each defect feature to obtain basic quality evaluation data of each defect feature; based on the basic quality evaluation data of each defect feature, combined with the overall defect distribution value and overall defect density value of the connector, performing overall quality evaluation processing on the connector to obtain comprehensive quality evaluation data of the connector; integrating the basic evaluation data of each defect feature of the connector and the comprehensive quality evaluation data of the connector to generate an inspection and evaluation data report for the connector.

[0012] The present invention also provides a connector detection device, comprising: a first scanning unit, used to use high-frequency ultrasonic detection equipment to perform an all-round scan of the connector to obtain an ultrasonic reflection image of the connector; a second scanning unit, used to use optical coherence tomography equipment to perform high-resolution surface and internal imaging of the connector to obtain an OCT three-dimensional image of the connector; an extraction unit, used to perform defect identification and feature extraction based on the ultrasonic reflection image and OCT three-dimensional image of the connector, to generate the defect feature list and detailed parameter information of each defect feature; an evaluation unit, used to perform detection and evaluation of the connector based on the defect feature list and the detailed parameter information of each defect feature to obtain a detection and evaluation data report of the connector.

[0013] The present invention also provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of any one of the above methods when executing the computer program.

[0014] The present invention also provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of any of the above methods are implemented.

[0015] The present invention provides a connector inspection method, apparatus, computer device, and storage medium. By combining high-frequency ultrasonic testing and optical coherence tomography (OCT) technology, this method provides a highly accurate and comprehensive connector inspection method. Specific contributions are as follows: The combination of high-frequency ultrasonic testing and OCT enables the inspection method to not only obtain surface information about the connector but also penetrate deep into its interior for comprehensive detection of internal defects. The high-resolution three-dimensional images provided by OCT accurately identify minute defects, improving inspection accuracy and reliability. Image processing and analysis algorithms can extract detailed defect parameters, providing more specific and useful inspection information. Comprehensive analysis of defect characteristics allows for a systematic assessment of connector quality, generating detailed inspection reports that support quality control and improvement. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 1 is a schematic diagram of the steps of a connector detection method according to an embodiment of the present invention;

[0017] Figure 2 is a structural block diagram of a connector detection device according to an embodiment of the present invention;

[0018] Figure 3 It is a schematic block diagram of the structure of a computer device according to an embodiment of the present invention.

[0019] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0020] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0021] Reference Figure 1 , an embodiment of the present invention provides a method for detecting a connector, comprising the following steps:

[0022] S1, using high-frequency ultrasonic detection equipment to perform an omnidirectional scan on the connector to obtain an ultrasonic reflection image of the connector.

[0023] This step generates an image of the connector's internal structure by transmitting high-frequency ultrasonic waves into the connector and receiving their reflected waves. Specifically, high-frequency ultrasonic waves penetrate the material and detect defects such as cracks and pores within the connector. Compared to traditional two-dimensional image testing, high-frequency ultrasonic testing can provide detailed information about the connector's internal structure and identify defects hidden beneath the surface, overcoming the limitation of two-dimensional images that only provide two-dimensional information.

[0024] S2, using an optical coherence tomography device to perform high-resolution surface and internal imaging of the connector to obtain an OCT three-dimensional image of the connector.

[0025] This step utilizes the coherence of light to obtain high-resolution, three-dimensional images of the interior of an object. By scanning the connector, images of its surface and internal fine structures can be captured. Specifically, OCT provides high-resolution, three-dimensional imaging capabilities, capable of displaying the connector's microscopic topography and internal details in detail. This complements the resolution and detail limitations of two-dimensional and ultrasonic images, thereby improving the comprehensiveness and accuracy of inspections.

[0026] S3, performing defect recognition and feature extraction based on the ultrasonic reflection image and the OCT three-dimensional image of the connector, and generating the defect feature list and detailed parameter information of each defect feature.

[0027] This step combines high-frequency ultrasound images with three-dimensional OCT images, using image processing and analysis algorithms to identify various defects in the connector (such as cracks, pores, and deformation) and extract detailed parameters for each defect (such as size, location, and morphology). Specifically, this step combines the advantages of two different imaging technologies to more comprehensively identify and analyze connector defects, providing detailed defect feature information and significantly improving the accuracy and comprehensiveness of defect detection.

[0028] S4, performing inspection and evaluation on the connector based on the defect feature list and detailed parameter information of each defect feature to obtain an inspection and evaluation data report of the connector.

[0029] This step comprehensively analyzes the identified and extracted defect characteristics and their detailed parameters to assess the overall quality and reliability of the connector and generate an inspection and evaluation report. This step provides a systematic assessment method that comprehensively evaluates connector quality, providing specific defect information and assessment results, helping manufacturers and quality inspectors make more accurate quality judgments and implement improvement measures.

[0030] Overall, this technical solution provides a highly accurate and comprehensive connector inspection method by combining high-frequency ultrasonic testing and optical coherence tomography (OCT). The specific contributions are as follows: 1. The combination of high-frequency ultrasonic testing and OCT enables the inspection method to not only obtain surface information about the connector, but also penetrate deep into the interior to comprehensively detect internal defects. 2. The high-resolution three-dimensional images provided by OCT can accurately identify tiny defects, improving the accuracy and reliability of inspection. 3. Through image processing and analysis algorithms, detailed defect parameters can be extracted, providing more specific and useful inspection information. 4. Comprehensive analysis of defect characteristics allows for a systematic assessment of connector quality, generating detailed inspection reports to support quality control and improvement.

[0031] It should be noted that quality inspection of micro-connectors is crucial in electronics manufacturing. Traditional 2D image inspection methods only see the connector surface and cannot detect potential internal defects. To address this issue, high-frequency ultrasound and optical coherence tomography (OCT) technologies are used for inspection.

[0032] First, a high-frequency ultrasonic inspection device performs a full-scale scan of the connector. The ultrasonic reflection image reveals tiny internal cracks and pores, defects that are invisible in traditional two-dimensional images. Next, an optical coherent imaging (OCT) device is used to perform high-resolution surface and internal imaging of the connector. The 3D OCT image reveals the connector's fine structure, revealing not only tiny surface flaws but also the more complex internal topography. This step provides a detailed view of every detail, both inside and out.

[0033] Combining ultrasound and OCT images, image processing algorithms identify and extract detailed parameters for each defect. For example, the length, depth, and location of cracks, as well as the size and distribution of pores, can be precisely measured. Finally, all defect characteristics and detailed parameters are comprehensively analyzed to generate a detailed inspection and evaluation report. This report lists defect information for each connector, assesses its overall quality, and provides improvement recommendations. This report ensures that connectors in electronic devices meet high quality and reliability requirements, avoiding potential failure risks. This approach significantly improves the accuracy and comprehensiveness of connector inspections, ensuring high product performance and stability.

[0034] In one example, a high-frequency ultrasonic detection device is used to perform an all-round scan of a connector to obtain an ultrasonic reflection image of the connector, including: scanning the connector with a high-frequency ultrasonic probe to collect initial ultrasonic reflection data from different angles and depths; performing denoising and signal enhancement processing on the initial ultrasonic reflection data to eliminate background noise in the initial ultrasonic reflection data and increase the signal strength of the initial ultrasonic reflection data to obtain preprocessed ultrasonic reflection data; performing a Hilbert transform on the preprocessed ultrasonic reflection data to obtain the imaginary part and the real part of a complex envelope signal, and calculating the envelope signal based on the imaginary part and the real part of the complex envelope signal; normalizing the envelope signal and converting the normalized envelope signal into an image format so that each value in the envelope signal is mapped to the grayscale or color value of the image to obtain an ultrasonic reflection image.

[0035] In this example, a high-frequency ultrasonic probe is used to scan the connector, collecting initial ultrasonic reflection data from various angles and depths. This omnidirectional scanning ensures that reflected signals from all parts of the connector are captured, providing comprehensive internal information. Next, the initial ultrasonic reflection data undergoes denoising and signal enhancement to eliminate background noise and increase signal strength. Denoising removes irrelevant background noise through filtering algorithms, while signal enhancement enhances the strength of the useful signal through methods such as gain adjustment. This step produces preprocessed ultrasonic reflection data, ensuring data clarity and accuracy. The preprocessed ultrasonic reflection data is then subjected to a Hilbert transform to obtain the imaginary and real parts of a complex envelope signal. The Hilbert transform is a mathematical process that converts a real signal into a complex signal, thereby calculating the envelope signal. The envelope signal reflects the variations in ultrasonic reflection intensity and can reveal the internal structural characteristics of the connector. The envelope signal is then calculated based on the imaginary and real parts of the complex envelope signal and normalized. Normalization brings the envelope signal within a standard range, facilitating subsequent image processing. Finally, the normalized envelope signal is converted into an image format, mapping each value in the envelope signal to the grayscale or color value of the image to produce an ultrasonic reflection image. This method visually displays changes in the intensity of the reflection signal as grayscale or color differences, helping to identify defects such as small cracks and pores within the connector.

[0036] For example, during the inspection of a micro-connector, a high-frequency ultrasonic probe scans from multiple angles, capturing subtle variations in the reflected signal. After noise removal and signal enhancement, the crack structure within the connector is clearly revealed. A Hilbert transform and normalization process accurately convert these variations into an image, clearly identifying the crack's location and shape. The resulting ultrasonic reflection image reveals the connector's detailed internal structure, providing reliable data for subsequent quality assessment and improvement. This example demonstrates the efficiency and accuracy of high-frequency ultrasonic testing in revealing internal defects in connectors.

[0037] In one example, an optical coherence tomography device is used to perform high-resolution surface and internal imaging of the connector to obtain an OCT three-dimensional image of the connector, including: scanning the connector using an optical coherence tomography scanner to obtain microstructural information on the surface and interior of the connector to obtain an original OCT data set; preprocessing the original OCT data set to sequentially eliminate background noise of the original OCT data set, correct depth signal attenuation of the original OCT data set, and correct phase error of the original OCT data set to obtain initial OCT data; using trilinear interpolation to calculate voxel values ​​of the initial OCT data, and based on the voxel values ​​of the initial OCT data, performing voxel reconstruction processing on the initial OCT data in combination with ray casting and maximum intensity projection to convert the two-dimensional data into three-dimensional data, and finally obtaining a reconstructed OCT three-dimensional image.

[0038] In other words, in electronics manufacturing, connector quality is crucial to the performance and stability of electronic devices. However, traditional 2D image inspection methods have significant limitations, providing only surface information on connectors and failing to fully capture internal defects and microstructures. To overcome these issues, this example solution employs optical coherence tomography (OCT) for high-resolution surface and internal imaging.

[0039] Specifically, an optical coherence tomography scanner is used to scan the connector to obtain its surface and internal microstructural information and obtain the original OCT data set. The technical effect of this step is that it can capture the subtle structure of the connector, including surface and internal information, providing a comprehensive foundation for subsequent data processing. Next, the original OCT data set is preprocessed, including eliminating background noise, correcting depth signal attenuation, and correcting phase errors to obtain initial OCT data. This step improves the quality and accuracy of the data by removing irrelevant noise and correcting errors in the data, making subsequent image reconstruction more reliable. Then, the voxel values ​​of the initial OCT data are calculated using trilinear interpolation. Based on these voxel values, the initial OCT data is voxel reconstructed using a combination of ray casting and maximum intensity projection techniques to convert the two-dimensional data into three-dimensional data. The technical effect of this process is that, through the calculation of voxel values ​​and three-dimensional reconstruction, complex two-dimensional data can be converted into intuitive three-dimensional images, fully displaying the internal and surface structures of the connector.

[0040] In summary, OCT technology captures high-resolution surface and internal microstructural information of connectors. Through complex preprocessing and 3D reconstruction steps, detailed OCT 3D images are ultimately generated. Compared to traditional 2D image inspection methods, this technical solution provides more comprehensive and accurate inspection results. Specifically, it overcomes the limitation of 2D image inspection, which only captures surface information, and can identify and analyze small defects and complex structures within connectors, significantly improving inspection accuracy and reliability. This comprehensive and accurate inspection method helps improve the overall quality and performance of electronic devices and reduces the risk of potential failures caused by connector issues.

[0041] In one example, defect identification and feature extraction are performed based on the ultrasonic reflection image of the connector to generate the defect feature list and detailed parameter information of each defect feature, including: removing high-frequency noise and low-frequency noise in the ultrasonic reflection image through a high-pass filter and a low-pass filter, and using a contrast enhancement algorithm to improve the clarity of the ultrasonic reflection image, thereby obtaining an initial ultrasonic reflection image in which the internal structure of the connector is more obvious; using a preset segmentation method based on region growing to segment the initial ultrasonic reflection image into multiple regions, wherein each region corresponds to a different component of the connector, and the components include at least: contact terminals and insulators; The cut image area is subjected to feature extraction to obtain internal and external morphological features and internal and external texture features of each component of the connector, and a machine learning algorithm based on a support vector machine is used to compare the extracted features with a preset connector defect feature library to identify the defect features represented by the initial ultrasonic reflection image and detailed parameter information of each defect feature; a defect feature list is generated based on the identified defect features, and the defect feature list is stored in a structured data format, and detailed parameter information of each defect feature is recorded in the defect feature list, wherein the detailed parameter information includes at least type data, position data, size data and severity information of the defect feature.

[0042] In other words, in electronics manufacturing, connector quality is crucial to the performance and stability of electronic devices. Traditional inspection methods rely primarily on two-dimensional image analysis, which presents significant challenges. For example, 2D images only provide information about the connector surface and fail to fully capture internal defects and microstructures. Furthermore, the lack of depth information and resolution makes accurate detection and analysis of minute defects and internal structures difficult. To address these issues, this example solution proposes a defect identification and feature extraction method based on ultrasonic reflection images.

[0043] Specifically, high-pass and low-pass filters are first used to remove high- and low-frequency noise from the ultrasonic reflection image, and a contrast enhancement algorithm is used to improve image clarity, resulting in an initial ultrasonic reflection image that more clearly highlights the connector's internal structure. The technical effect of this step is to significantly improve image clarity, making the connector's internal structure more prominent and easier to identify. Next, a preset segmentation method based on region growing is used to segment the initial ultrasonic reflection image into multiple regions, each corresponding to a different connector component, including the contact terminals and insulator. This segmentation method accurately distinguishes the various connector components, laying the foundation for subsequent feature extraction and defect identification. Feature extraction is then performed on the segmented image regions to obtain internal and external morphological and texture features of each connector component. Using a machine learning algorithm based on a support vector machine, the extracted features are compared with a preset connector defect feature library to identify the defect features represented in the initial ultrasonic reflection image, as well as detailed parameter information for each defect feature. The technical effect of this step is that the machine learning algorithm can accurately identify various defect features and provide detailed parameter information, including defect type, location, size, and severity. Finally, a defect feature list is generated based on the identified defect features and stored in a structured data format. The detailed parameter information for each defect feature is also recorded in the feature list. This allows for systematic recording and management of defect information, facilitating subsequent quality assessment and improvement.

[0044] In summary, the preprocessing and contrast enhancement of ultrasonic reflection images significantly improved image clarity, making the internal structure of the connector more prominent. The region growing-based segmentation method divides the image into multiple regions, accurately locating the various components of the connector. The combination of feature extraction and machine learning algorithms makes defect identification more accurate and detailed, providing comprehensive information such as the type, location, size, and severity of the defect. The final defect feature list is stored in a structured data format to facilitate systematic management and subsequent analysis. Overall, this technical solution significantly improves the comprehensiveness and accuracy of detection, compensates for the shortcomings of traditional two-dimensional image detection, and can more comprehensively reflect the internal defects and microstructure of the connector, improving product quality and reliability.

[0045] In one example, defect recognition and feature extraction are performed based on the OCT three-dimensional image of the connector to generate the defect feature list and detailed parameter information of each defect feature, including: removing high-frequency noise and low-frequency noise in the OCT three-dimensional image by an adaptive stop filtering method, and using a histogram equalization method to improve the image contrast of the OCT three-dimensional image to obtain an initial OCT three-dimensional image representing the internal and external structure of the connector; using a preset segmentation method based on region growing to segment the initial OCT three-dimensional image into multiple regions, wherein each region corresponds to a different component of the connector, and the components include at least: contact terminals and insulators; The method comprises the following steps: extracting features from the image area to obtain the internal and external morphological features and internal and external texture features of each component of the connector, and using a machine learning algorithm based on a support vector machine to compare the extracted features with a preset connector defect feature library to identify the defect features of the connector represented by the initial OCT three-dimensional image, as well as detailed parameter information of each defect feature; generating a defect feature list based on the identified defect features, storing the defect feature list in a structured data format, and recording the detailed parameter information of each defect feature in the defect feature list, wherein the detailed parameter information includes at least type data, position data, size data and severity information of the defect feature.

[0046] In other words, in electronics manufacturing, ensuring connector quality is crucial to the performance and stability of electronic devices. However, traditional 2D image inspection methods have significant limitations. They only provide information on the connector surface and cannot fully reflect internal defects and microstructures. To address these issues, a method for defect identification and feature extraction based on 3D optical coherence tomography (OCT) images was proposed.

[0047] Specifically, the method uses high-resolution OCT 3D imaging to obtain surface and internal microstructural information. First, adaptive stop filtering is used to remove high- and low-frequency noise from the OCT 3D image, and histogram equalization is used to enhance image contrast, resulting in an initial OCT 3D image depicting the connector's internal and external structures. This step significantly improves image clarity and contrast, making the connector's structure more distinct and easier to analyze. Next, a pre-defined segmentation method based on region growing is used to segment the initial OCT 3D image into multiple regions, each corresponding to a different connector component, including the contact terminals and insulator. This segmentation method accurately distinguishes the various connector components, laying the foundation for subsequent feature extraction and defect identification. Feature extraction is then performed on the segmented image regions to obtain internal and external morphological and texture features of each connector component. Using a machine learning algorithm based on a support vector machine (SVM), the extracted features are compared with a pre-defined connector defect feature library to identify connector defect features in the initial OCT 3D image and extract detailed parameter information for each defect feature. This step uses machine learning algorithms to accurately identify various defect characteristics and provide detailed parameter information, including defect type, location, size, and severity. Finally, a defect feature list is generated based on the identified defect characteristics and stored in a structured data format. The detailed parameter information for each defect feature is also recorded in the feature list. This method systematically records and manages defect information, facilitating subsequent quality assessment and improvement.

[0048] In summary, the entire technical solution significantly improves the clarity and contrast of the image through preprocessing and contrast enhancement of OCT three-dimensional images, making the internal and external structures of the connector more prominent. The segmentation method based on region growing divides the image into multiple regions to accurately locate the various components of the connector. The combination of feature extraction and machine learning algorithms makes defect identification more accurate and detailed, and can provide comprehensive information such as the type, location, size and severity of the defect. The final defect feature list is stored in a structured data format to facilitate systematic management and subsequent analysis. Overall, this technical solution significantly improves the comprehensiveness and accuracy of detection, makes up for the shortcomings of traditional two-dimensional image detection, and can more comprehensively reflect the internal defects and microstructure of the connector, thereby improving product quality and reliability.

[0049] In one example, the connector is inspected and evaluated based on the defect feature list and detailed parameter information of each defect feature to obtain an inspection and evaluation data report of the connector, including: classifying the defect features of the connector according to the defect type field in the defect feature list, and extracting the position data, size data and severity information of each defect feature in each type of defect feature in turn; calculating the area ratio value and defect depth value of each defect feature, as well as the overall defect distribution value and overall defect density value of the connector based on the position data and size data of each defect feature; determining the weight distribution value corresponding to each defect feature based on the severity information of each defect feature, and performing weighted calculation on the area ratio value and defect depth value of each defect feature based on the weight distribution value corresponding to each defect feature to obtain basic quality assessment data of each defect feature; performing overall quality assessment processing on the connector based on the basic quality assessment data of each defect feature, combined with the overall defect distribution value and overall defect density value of the connector, to obtain comprehensive quality assessment data of the connector; integrating the basic evaluation data of each defect feature of the connector and the comprehensive quality assessment data of the connector to generate an inspection and evaluation data report for the connector.

[0050] In other words, in electronics manufacturing, connector quality is crucial to the performance and stability of equipment. Traditional inspection methods rely primarily on two-dimensional image analysis, which suffers from incomplete information and insufficient accuracy. To address these issues, a connector inspection and evaluation method based on a defect signature list and detailed parameter information is proposed to generate an inspection and evaluation data report.

[0051] Specifically, the connector's defect features are first classified according to the defect type field in the defect feature list. The location, size, and severity information for each defect feature within each category is then extracted. This step systematically organizes and categorizes defect information, making subsequent data analysis more organized and efficient. Next, based on the location and size data for each defect feature, the area percentage and defect depth of each defect feature, as well as the connector's overall defect distribution and overall defect density, are calculated. This step quantifies the defect features, providing a detailed picture of the defect percentage and distribution, making the inspection results more specific and comparable. Subsequently, based on the severity information of each defect feature, a weighted value is assigned to each defect feature. The area percentage and defect depth values ​​of each defect feature are then weighted to obtain basic quality assessment data for each defect feature. This process, by introducing a weighting mechanism and combining defect severity, provides a more comprehensive and accurate defect assessment. Finally, based on the basic quality assessment data for each defect feature, combined with the connector's overall defect distribution and overall defect density, a comprehensive connector quality assessment is performed, resulting in comprehensive quality assessment data for the connector. This step integrates individual defects and their overall distribution, providing a comprehensive assessment of the connector's overall quality. Finally, the basic evaluation data for each defect characteristic of the connector, along with the connector's comprehensive quality assessment data, is integrated to generate a connector inspection and evaluation data report. This report comprehensively records the detailed information for each defect characteristic and the overall quality assessment results, providing a systematic and structured inspection report.

[0052] In summary, the entire technical solution systematically organizes defect information by classifying defect characteristics and extracting detailed parameters. It quantifies the area percentage and depth of each defect and, combined with severity information, performs a weighted calculation, providing accurate basic quality assessment data. Combining the overall defect distribution and density, it conducts a comprehensive connector quality assessment and generates a comprehensive inspection and evaluation data report. This technical solution significantly improves the comprehensiveness and accuracy of inspections, addressing the shortcomings of traditional methods, providing a more specific, systematic, and actionable quality assessment, and enhancing the effectiveness and reliability of connector inspections.

[0053] In addition, the present invention also provides another parallel solution to solve the problem that traditional detection methods mainly rely on two-dimensional image analysis, which has incomplete information and insufficient accuracy. The details are as follows:

[0054] In one example, defect recognition and feature extraction are performed based on the ultrasonic reflection image and the OCT three-dimensional image of the connector to generate the defect feature list and detailed parameter information of each defect feature, including: using a fusion algorithm based on feature point matching and image registration to spatially fuse the ultrasonic reflection image and the OCT three-dimensional image of the connector to obtain a fused target three-dimensional image; defect recognition and feature extraction are performed on the fused target three-dimensional image to generate the defect feature list and detailed parameter information of each defect feature.

[0055] It's important to note that the benefit of fusing ultrasonic reflection images with OCT 3D images lies in the complementary strengths of each imaging technique. Ultrasonic reflection images reveal the connector's internal structure, enabling in-depth detection of hidden defects. OCT 3D images, on the other hand, offer high resolution, clearly showcasing the connector's surface and microstructure. By fusing these two images, detailed information on both the connector's interior and surface can be obtained simultaneously, providing a more comprehensive inspection foundation.

[0056] However, the image fusion process also faces certain difficulties. First, the feature points of the ultrasound reflection image and the OCT three-dimensional image need to be accurately matched, which requires complex feature point extraction and matching algorithms. Second, images from different imaging technologies need to be precisely registered spatially to ensure that the data of the two images accurately correspond. Finally, the fused image needs to retain the key information of both the ultrasound and OCT images, which requires an efficient image fusion algorithm to ensure the quality and accuracy of the final fused image.

[0057] In one example, a fusion algorithm based on feature point matching and image registration is used to spatially fuse the ultrasonic reflection image and the OCT three-dimensional image of the connector to obtain a fused target three-dimensional image, including: extracting feature points of the ultrasonic reflection image using a first formula; extracting feature points of the OCT three-dimensional image using a second formula; matching the feature points of the ultrasonic reflection image with the feature points of the OCT three-dimensional image using a feature matching algorithm to obtain corresponding feature point pairs of the ultrasonic reflection image and the OCT three-dimensional image; calculating and processing the corresponding feature point pairs of the ultrasonic reflection image and the OCT three-dimensional image using a preliminary registration matrix; transforming points in the ultrasonic reflection image to corresponding positions in the OCT three-dimensional image based on the calculation results to obtain registered image data of the ultrasonic reflection image and the OCT three-dimensional image; and using an image fusion algorithm based on wavelet transform to perform image fusion processing on the registered image data to obtain the target three-dimensional image, wherein the target three-dimensional image includes: internal structure information of the ultrasonic reflection image and surface structure information and microstructure information of the OCT three-dimensional image.

[0058] It should be noted that the first formula is: S(u, v) = (∂I / ∂x)^2 + (∂I / ∂y)^2, where S(u, v) is the characteristic point intensity of the ultrasound reflection image, I represents the image intensity, and ∂I / ∂x and ∂I / ∂y are the image gradients in the x and y directions, respectively. The second formula is; T(x, y, z) = √((∂V / ∂x)^2 + (∂V / ∂y)^2 + (∂V / ∂z)^2), where T(x, y, z) is the characteristic point intensity of the OCT 3D image, V represents the voxel intensity, and ∂V / ∂x, ∂V / ∂y, and ∂V / ∂z are the voxel gradients in the x, y, and z directions, respectively.

[0059] In addition, the image fusion formula in the wavelet transform-based image fusion algorithm is: F(x, y, z)=W(U(x, y),O(x, y, z)), where F(x, y, z) represents the fused target three-dimensional image, W is the wavelet transform function, U(x, y) represents the intensity information of the registered ultrasonic reflection image, and O(x, y, z) represents the intensity information of the registered OCT three-dimensional image.

[0060] In this example, the first formula is used to extract the feature points of the ultrasound reflection image. This formula is: S(u,v) = (∂I / ∂x)^2 + (∂I / ∂y)^2, where S(u, v) represents the intensity of the feature points in the ultrasound reflection image, I represents the image intensity, and ∂I / ∂x and ∂I / ∂y represent the image gradients in the x and y directions, respectively. This step enhances the characteristic information of the ultrasound image by extracting edges and significant feature points. Next, the second formula is used to extract feature points from the 3D OCT image. The formula is: T(x, y, z) = √((∂V / ∂x)^2 + (∂V / ∂y)^2 + (∂V / ∂z)^2), where T(x, y, z) is the feature point intensity of the OCT 3D image, V represents the voxel intensity, and ∂V / ∂x, ∂V / ∂y, and ∂V / ∂z are the gradients of the voxel in the x, y, and z directions, respectively. This step makes the feature information of the OCT image clearer by extracting the significant feature points of the 3D image. Then, a feature matching algorithm is used to match the feature points of the ultrasound reflection image with the feature points of the OCT 3D image to obtain corresponding feature point pairs of the ultrasound reflection image and the OCT 3D image. This step achieves spatial alignment between the images by matching the feature points of the two images, providing a basis for subsequent registration and fusion. Next, a preliminary registration matrix is ​​used to calculate and process corresponding feature point pairs between the ultrasound reflection image and the OCT 3D image. Based on the calculated results, points in the ultrasound reflection image are transformed to corresponding locations in the OCT 3D image to obtain the registered image data. This ensures precise spatial alignment of the two images, ensuring accurate correspondence between the two data. Finally, a wavelet-based image fusion algorithm is used to fuse the registered image data to obtain the target 3D image. The wavelet transform formula is: F(x, y, z) = W(U(x, y), O(x, y, z)), where F(x, y, z) represents the fused target 3D image, W is the wavelet transform function, U(x, y) represents the intensity information of the registered ultrasound reflection image, and O(x, y, z) represents the intensity information of the registered OCT 3D image. Image fusion using wavelet transform effectively combines the advantages of both images, obtaining a target 3D image that contains both internal structural information from the ultrasound reflection image and surface and microstructural information from the OCT 3D image.

[0061] Overall, the entire technical solution fuses ultrasonic reflection images with OCT 3D images, combining the strengths of both imaging technologies to generate a target 3D image that comprehensively reflects the connector's internal and external structure. This process utilizes steps such as feature point extraction and matching, preliminary registration matrix calculation, and wavelet transform image fusion to ensure image quality and accuracy. This technical solution significantly improves the comprehensiveness and accuracy of inspection, addressing the shortcomings of traditional 2D image inspection and enabling a more comprehensive understanding of connector internal defects and microstructure, thereby enhancing product quality and reliability.

[0062] In one example, defect recognition and feature extraction are performed on the fused target three-dimensional image to generate the defect feature list and parameter information of each defect feature, including: using an adaptive stop filtering method to remove high-frequency noise in the target three-dimensional image, and using a low-pass filter to eliminate low-frequency noise, to obtain the denoised target three-dimensional image; using a histogram equalization method to improve the overall contrast of the denoised target three-dimensional image, so that the structural features in the image are more obvious, to obtain the target three-dimensional image after image enhancement; performing edge enhancement processing on the target three-dimensional image after image enhancement by using a Laplace operator to enhance the edge information in the target three-dimensional image and enhance the structural details in the target three-dimensional image, to obtain the target three-dimensional image after edge enhancement processing; using a segmentation method based on region growing to segment the target three-dimensional image after edge enhancement processing into multiple regions, each region corresponding to a different component of the connector, the components including at least: contact terminals and insulators; performing morphological feature extraction and texture feature extraction on the multiple segmented regions to obtain the connector The internal and external morphological features and internal and external texture features of each component of the connector are classified by a machine learning algorithm based on a support vector machine (SVM), and compared with a preset connector defect feature library to identify defect features in the target three-dimensional image and determine parameter information of the defect features, wherein the parameter information includes at least defect type, defect size and defect severity; based on the voxel information of the target three-dimensional image, the three-dimensional morphological features and spatial distribution features of the multiple regions are determined, and the position information of each defect feature is determined by combining the three-dimensional morphological features and spatial distribution features of each region; the identified defect features are recorded in the form of a structured data table to generate a defect feature list, and the parameter information and position information corresponding to each defect feature are recorded in the defect feature list.

[0063] In this example, an adaptive stop filter is first used to remove high-frequency noise from the target 3D image, and a low-pass filter is used to remove low-frequency noise, resulting in a denoised target 3D image. This denoising process effectively eliminates both high- and low-frequency noise from the image, making the image clearer and providing high-quality data for subsequent processing. Next, histogram equalization is used to enhance the overall contrast of the denoised target 3D image, making structural features more distinct. This results in an enhanced target 3D image. Histogram equalization makes the image's brightness distribution more uniform, enhancing contrast and highlighting details, laying the foundation for subsequent edge enhancement. Finally, edge enhancement is performed on the enhanced target 3D image using the Laplacian operator, enhancing edge information and structural details. This step enhances edges and makes structures clearer and sharper, facilitating subsequent image segmentation and feature extraction.

[0064] In the next step, a region growing-based segmentation method is used to segment the edge-enhanced target 3D image into multiple regions, each corresponding to a different component of the connector, including the contact terminals and insulator. By selecting seed points and expanding the region, the region growing method can accurately segment different component regions, making the boundaries of each component clearly visible. Morphological and texture features are extracted from the multiple segmented regions to obtain the internal and external morphological features and internal and external texture features of each connector component. Internal and external morphological features include shape features, size features, and geometric structure features, while internal and external texture features include contrast features, entropy features, uniformity features, and correlation features. By extracting these features, the geometric shape and texture distribution of each region can be described in detail, providing rich feature data for defect identification.

[0065] Then, a machine learning algorithm based on a support vector machine (SVM) classifies the internal and external morphological and texture features of each connector component and compares them with a pre-defined connector defect signature library. Defect features within the target 3D image are identified and parameter information for each defect feature is determined. This parameter information includes at least defect type, size, and severity. By classifying and comparing feature data, the SVM algorithm efficiently and accurately identifies various defect types. Based on the voxel information within the target 3D image, the 3D morphological and spatial distribution features of multiple regions are determined. Combining these features, the location of each defect feature is determined. By analyzing the 3D voxel data, the spatial location of each defect can be precisely determined, providing a crucial basis for comprehensive quality assessment. Finally, the identified defect features are recorded in a structured data table to generate a defect feature list. The parameter information and location information for each defect feature are recorded in the defect feature list. This systematic recording and management of defect information facilitates subsequent quality assessment and improvement.

[0066] Overall, this technical solution, through a series of image processing and feature extraction steps, combined with machine learning algorithms, can comprehensively and accurately identify and record defect signatures in connectors. Through steps such as denoising, contrast enhancement, edge enhancement, image segmentation, feature extraction, and classification, comprehensiveness and accuracy of inspection are ensured. The resulting defect signature list provides detailed and reliable data support for quality assessment, significantly improving the effectiveness and reliability of connector inspection.

[0067] Reference Figure 2 , an embodiment of the present invention provides a connector detection device, comprising:

[0068] The first scanning unit 1 is used to perform an omnidirectional scan of the connector using a high-frequency ultrasonic detection device to obtain an ultrasonic reflection image of the connector;

[0069] A second scanning unit 2 is configured to perform high-resolution surface and internal imaging of the connector using an optical coherence tomography device to obtain an OCT three-dimensional image of the connector;

[0070] Extraction unit 3, used to perform defect identification and feature extraction based on the ultrasonic reflection image and OCT three-dimensional image of the connector, and generate the defect feature list and detailed parameter information of each defect feature;

[0071] The evaluation unit 4 is configured to perform detection and evaluation on the connector based on the defect feature list and detailed parameter information of each defect feature, and obtain a detection and evaluation data report of the connector.

[0072] In this embodiment, for the specific implementation of each unit in the above device embodiment, please refer to the above method embodiment, which will not be repeated here.

[0073] Reference Figure 3 In an embodiment of the present invention, a computer device is also provided. The computer device may be a server, and its internal structure may be as follows: Figure 3 As shown. The computer device includes a processor, memory, display screen, input device, network interface and database connected via a system bus. The processor of the computer design is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The database of the computer device is used to store the corresponding data in this embodiment. The network interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, the above method is implemented.

[0074] Those skilled in the art will understand that Figure 3 The structure shown in the figure is merely a block diagram of a portion of the structure related to the solution of the present invention and does not constitute a limitation on the computer device to which the solution of the present invention is applied.

[0075] An embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon, which implements the above-described method when executed by a processor. It is understood that the computer-readable storage medium in this embodiment can be a volatile readable storage medium or a non-volatile readable storage medium.

[0076] Those skilled in the art will appreciate that all or part of the processes in the above-described method embodiments can be implemented by instructing the relevant hardware using a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the above-described method embodiments. Any reference to memory, storage, database, or other media provided herein and used in the embodiments may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double-speed SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct RAMbus dynamic RAM (DRDRAM), and RAMbus dynamic RAM.

[0077] It should be noted that, in this document, the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, apparatus, article, or method comprising a series of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, apparatus, article, or method. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, apparatus, article, or method comprising the element.

[0078] The above description is only a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the contents of the present invention description and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for detecting a connector, characterized in that: The following steps are involved: Using high-frequency ultrasonic testing equipment to perform an omnidirectional scan of the connector to obtain an ultrasonic reflection image of the connector; Using an optical coherence tomography device to perform high-resolution surface and internal imaging of the connector to obtain an OCT three-dimensional image of the connector; Defect identification and feature extraction are performed based on the ultrasonic reflection image and the OCT three-dimensional image of the connector to generate a defect feature list and detailed parameter information of each defect feature; specifically, the method includes: removing high-frequency noise and low-frequency noise in the ultrasonic reflection image by using a high-pass filter and a low-pass filter, and improving the clarity of the ultrasonic reflection image by using a contrast enhancement algorithm to obtain an initial ultrasonic reflection image with a more obvious internal structure of the connector; using a preset segmentation method based on region growing to segment the initial ultrasonic reflection image into multiple regions, wherein each region corresponds to a different component of the connector, and the components include at least: contact terminals and insulators; Performing feature extraction on the segmented image area to obtain internal and external morphological features and internal and external texture features of each component of the connector, and using a machine learning algorithm based on a support vector machine to compare the extracted features with a preset connector defect feature library to identify defect features represented by the initial ultrasonic reflection image and detailed parameter information of each defect feature; generating a defect feature list based on the identified defect features, storing the defect feature list in a structured data format, and recording detailed parameter information of each defect feature in the defect feature list, wherein the detailed parameter information includes at least type data, location data, size data, and severity information of the defect feature; The connector is inspected and evaluated based on the defect feature list and detailed parameter information of each defect feature to obtain an inspection and evaluation data report of the connector.

2. The detection method according to claim 1, characterized in that A high-frequency ultrasonic testing device is used to perform an omnidirectional scan of the connector to obtain an ultrasonic reflection image of the connector, including: Scanning the connector using a high-frequency ultrasonic probe to collect initial ultrasonic reflection data from different angles and depths; performing denoising and signal enhancement processing on the initial ultrasonic reflection data to eliminate background noise in the initial ultrasonic reflection data and enhance the signal strength of the initial ultrasonic reflection data to obtain preprocessed ultrasonic reflection data; performing a Hilbert transform on the preprocessed ultrasonic reflection data to obtain an imaginary part and a real part of a complex envelope signal, and calculating the envelope signal based on the imaginary part and the real part of the complex envelope signal; The envelope signal is normalized and converted into an image format so that each value in the envelope signal is mapped to a grayscale or color value of an image to obtain an ultrasonic reflection image.

3. The detection method according to claim 1, wherein An optical coherence tomography device is used to perform high-resolution surface and internal imaging of the connector to obtain an OCT three-dimensional image of the connector, including: Scanning the connector using an optical coherence tomography scanner to obtain microstructural information on the surface and interior of the connector, thereby obtaining an original OCT dataset; Preprocessing the original OCT data set, sequentially eliminating background noise of the original OCT data set, correcting depth signal attenuation of the original OCT data set, and correcting phase error of the original OCT data set to obtain initial OCT data; The voxel values ​​of the initial OCT data are calculated using a trilinear interpolation method, and based on the voxel values ​​of the initial OCT data, the initial OCT data is subjected to voxel reconstruction processing in combination with the ray casting and maximum intensity projection, the two-dimensional data is converted into three-dimensional data, and finally a reconstructed OCT three-dimensional image is obtained.

4. The detection method according to claim 3, characterized in that Defect identification and feature extraction are performed based on the OCT three-dimensional image of the connector to generate a defect feature list and detailed parameter information of each defect feature, including: removing high-frequency noise and low-frequency noise from the OCT three-dimensional image by an adaptive stop filtering method, and improving the image contrast of the OCT three-dimensional image by a histogram equalization method, thereby obtaining an initial OCT three-dimensional image representing the internal and external structure of the connecting machine; Using a preset segmentation method based on region growing, the initial OCT three-dimensional image is segmented into a plurality of regions, wherein each region corresponds to a different component of the connector, the components including at least: a contact terminal and an insulator; Performing feature extraction on the segmented image area to obtain internal and external morphological features and internal and external texture features of each component of the connector, and using a machine learning algorithm based on a support vector machine to compare the extracted features with a preset connector defect feature library to identify defect features present in the connector as represented by the initial OCT three-dimensional image, as well as detailed parameter information for each defect feature; A defect feature list is generated based on the identified defect features, and the defect feature list is stored in a structured data format, and detailed parameter information of each defect feature is recorded in the defect feature list, wherein the detailed parameter information includes at least type data, location data, size data and severity information of the defect feature.

5. The detection method according to claim 4, characterized in that The connector is inspected and evaluated based on the defect feature list and detailed parameter information of each defect feature to obtain an inspection and evaluation data report of the connector, including: Classifying the defect features of the connector according to the defect type field in the defect feature list, and sequentially extracting position data, size data, and severity information of each defect feature in each category of defect features; Based on the position data and size data of each defect feature, calculating the area ratio value and defect depth value of each defect feature, as well as the overall defect distribution value and overall defect density value of the connector; Determining a weight value corresponding to each defect feature based on the severity information of each defect feature, and performing a weighted calculation on the area ratio value and the defect depth value of each defect feature based on the weight value corresponding to each defect feature to obtain basic quality assessment data for each defect feature; Based on the basic quality assessment data of each defect feature, combined with the overall defect distribution value and the overall defect density value of the connector, an overall quality assessment process is performed on the connector to obtain comprehensive quality assessment data of the connector; The basic evaluation data of each defect feature of the connector and the comprehensive quality evaluation data of the connector are integrated to generate an inspection and evaluation data report of the connector.

6. A connector detection device, characterized in that: include: A first scanning unit is used to perform an omnidirectional scan of the connector using a high-frequency ultrasonic detection device to obtain an ultrasonic reflection image of the connector; a second scanning unit, configured to perform high-resolution surface and internal imaging of the connector using an optical coherence tomography device to obtain an OCT three-dimensional image of the connector; An extraction unit is configured to perform defect recognition and feature extraction based on the ultrasonic reflection image and the OCT three-dimensional image of the connector, and generate a defect feature list and detailed parameter information of each defect feature; specifically comprising: removing high-frequency noise and low-frequency noise in the ultrasonic reflection image by a high-pass filter and a low-pass filter, and improving the clarity of the ultrasonic reflection image by a contrast enhancement algorithm, so as to obtain an initial ultrasonic reflection image with a more obvious internal structure of the connector; and using a preset segmentation method based on region growing to segment the initial ultrasonic reflection image into a plurality of regions, wherein each region corresponds to a different component of the connector, and the components include at least: contact terminals and insulation. edge body; performing feature extraction on the segmented image area to obtain internal and external morphological features and internal and external texture features of each component of the connector, and using a machine learning algorithm based on a support vector machine to compare the extracted features with a preset connector defect feature library to identify the defect features represented by the initial ultrasonic reflection image and detailed parameter information of each defect feature; generating a defect feature list based on the identified defect features, storing the defect feature list in a structured data format, and recording detailed parameter information of each defect feature in the defect feature list, wherein the detailed parameter information includes at least type data, location data, size data, and severity information of the defect feature; An evaluation unit is configured to perform detection and evaluation on the connector based on the defect feature list and detailed parameter information of each defect feature, and obtain a detection and evaluation data report of the connector.

7. A computer device comprising a memory and a processor, wherein a computer program is stored in the memory, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 5 are implemented.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 5 are implemented.

Images