A defect identification method, device, system and computer equipment

Abstract

This application relates to a defect identification method, apparatus, system, and computer device. It includes: determining the hardware parameters of a data acquisition device; obtaining a capability assessment result of the data acquisition device based on the hardware parameters; determining a defect identification algorithm matched with the data acquisition device based on a preset matching relationship and the capability assessment result; performing data acquisition and processing on a preset object to be identified by the data acquisition device to obtain data to be identified; and performing defect identification processing on the data to be identified based on the defect identification algorithm to obtain a defect identification result. This application can improve the accuracy of defect identification.

Description

Technical Field

[0001] This application relates to the field of defect identification technology, and in particular to a defect identification method, apparatus, system, and computer equipment. Background Technology

[0002] Defect identification refers to the process of capturing characteristic information of a target object through various detection devices, thereby determining whether the target object has defects, as well as the type and level of the defects.

[0003] In related technologies, fixed defect identification algorithms and fixed detection hardware configurations are often used to identify defects in target objects. That is, the defect identification algorithm and parameters are preset and matched with a fixed model of detection equipment. However, in actual applications, the defect identification algorithm and the detection equipment are incompatible. For example, when the detection equipment cannot capture the key features of a certain type of defect, the preset and fixed defect identification algorithm is prone to missed detections and false detections. Furthermore, hardware equipment will experience aging and wear during long-term operation, which will lead to a decrease in the accuracy of its data collection. However, the defect identification should not be adjusted in time, and it is impossible to compensate for the negative impact of hardware performance degradation in a timely manner.

[0004] Currently, no effective solution has been proposed to address the issue of low accuracy in defect identification in related technologies. Summary of the Invention

[0005] Therefore, it is necessary to provide a defect identification method, apparatus, system, or computer equipment to address the aforementioned technical problems.

[0006] Firstly, this application provides a defect identification method, including:

[0007] Determine the hardware parameters of the data acquisition equipment, and obtain the capability assessment results of the data acquisition equipment based on the hardware parameters;

[0008] Based on the preset matching relationship and the capability assessment results, a defect identification algorithm that matches the data acquisition equipment is determined.

[0009] The data acquisition device collects and processes data from the preset object to be identified to obtain the data to be identified. Based on the defect identification algorithm, the data to be identified is processed to obtain the defect identification result.

[0010] In one embodiment, the data acquisition device performs data acquisition processing on a preset object to be identified to obtain data to be identified. Based on a defect identification algorithm, the data to be identified is then processed for defect identification to obtain a defect identification result, including:

[0011] The working parameters of the data acquisition equipment are calibrated based on the defect identification algorithm, and the adjustable parameters in the defect identification algorithm are adjusted according to the hardware parameters of the data acquisition equipment.

[0012] The data acquisition equipment is used to collect and process data of the object to be identified after calibration, and the data to be identified is obtained. Based on the adjusted defect identification algorithm, the defect identification data is processed to obtain the defect identification result.

[0013] In one embodiment, the method further includes:

[0014] After obtaining the defect identification results, the performance status of the data acquisition equipment is tested;

[0015] When the performance status of the data acquisition device is detected to match the preset calibration conditions, new hardware parameters of the data acquisition device are collected, new capability assessment results of the data acquisition device are determined based on the new hardware parameters, and a new defect identification algorithm is determined based on the new capability assessment results.

[0016] In one embodiment, the hardware parameters of the data acquisition device are determined, and a capability assessment result of the acquisition device is obtained based on the hardware parameters, including:

[0017] The initial hardware parameters of the data acquisition device are collected, and the initial hardware parameters are normalized to obtain the hardware parameters; the hardware parameters include multiple parameter levels.

[0018] Determine the weights corresponding to each parameter level, and calculate the initial capability assessment results based on the hardware parameters and their corresponding weights.

[0019] The stability and key performance indicators of the data acquisition equipment are obtained. Based on the initial capability assessment results, the stability and key performance indicators, the capability assessment results are obtained.

[0020] In one embodiment, a defect identification algorithm matching the data acquisition device is determined based on a preset matching relationship and capability assessment results, including:

[0021] Obtain the matching relationship between multiple preset defect identification algorithms and the hardware capability range;

[0022] The target hardware capability range is determined from multiple hardware capability ranges based on the capability assessment results;

[0023] The corresponding defect identification algorithm is determined based on the target hardware capability range.

[0024] In one embodiment, the method further includes:

[0025] Determine the confidence level corresponding to each defect identification result, and count the target confidence levels among all confidence levels that are greater than the preset confidence threshold;

[0026] When the percentage of the target confidence level among all confidence levels does not reach the preset standard threshold, at least one of the following adjustment schemes shall be executed: adjust the defect identification algorithm that matches the data acquisition device, calibrate the working parameters of the data acquisition device, or adjust the adjustable parameters of the defect identification algorithm.

[0027] Based on the adjusted data acquisition equipment and defect identification algorithm, the object to be identified is re-detected, and the above steps are repeated until the target confidence level accounts for a certain percentage of all confidence levels and reaches the standard threshold.

[0028] Secondly, this application also provides a defect identification device, comprising:

[0029] The determination module is used to determine the hardware parameters of the data acquisition device and obtain the capability assessment results of the acquisition device based on the hardware parameters.

[0030] The matching module is used to determine the defect identification algorithm that matches the data acquisition equipment based on the preset matching relationship and the capability assessment results.

[0031] The generation module is used to collect and process data of the preset object to be identified by the data acquisition device to obtain the data to be identified, and to perform defect identification processing on the data to be identified based on the defect identification algorithm to obtain the defect identification result.

[0032] Thirdly, this application also provides a defect identification system, including:

[0033] The hardware acquisition module is used to determine the hardware parameters of the data acquisition device;

[0034] The capability assessment module is used to obtain the capability assessment results of the acquisition device based on the hardware parameters;

[0035] The algorithm matching module is used to determine the defect identification algorithm that matches the data acquisition equipment based on the preset matching relationship and the capability assessment results.

[0036] The defect identification module is used to collect and process data from a preset object to be identified using a data acquisition device to obtain the data to be identified. Based on the defect identification algorithm, the module performs defect identification processing on the data to be identified to obtain the defect identification result.

[0037] In one embodiment, the system further includes a collaborative calibration module and a result optimization module;

[0038] The collaborative calibration module is used to calibrate the working parameters of the data acquisition equipment based on the defect identification algorithm, and to adjust the adjustable parameters in the defect identification algorithm according to the hardware parameters of the data acquisition equipment.

[0039] The result optimization module is used to collect new hardware parameters of the data acquisition device when the performance status of the data acquisition device is detected to match the preset calibration conditions, determine new capability assessment results of the data acquisition device based on the new hardware parameters, and determine a new matching defect identification algorithm based on the new capability assessment results.

[0040] Fourthly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0041] Determine the hardware parameters of the data acquisition equipment, and obtain the capability assessment results of the data acquisition equipment based on the hardware parameters;

[0042] Based on the preset matching relationship and the capability assessment results, a defect identification algorithm that matches the data acquisition equipment is determined.

[0043] The data acquisition device collects and processes data from the preset object to be identified to obtain the data to be identified. Based on the defect identification algorithm, the data to be identified is processed to obtain the defect identification result.

[0044] The aforementioned defect identification method, apparatus, system, and computer equipment determine the hardware parameters of a data acquisition device and obtain a capability assessment result of the data acquisition device based on the hardware parameters. Based on a preset matching relationship, a defect identification algorithm matching the data acquisition device is determined according to the capability assessment result. The data acquisition device performs data acquisition and processing on a preset object to be identified to obtain data to be identified. Based on the defect identification algorithm, defect identification processing is performed on the data to be identified to obtain a defect identification result. In this embodiment, by collecting multi-dimensional hardware parameters and performing hardware capability assessment, precise matching between the defect identification algorithm and hardware capabilities is achieved. This allows for the selection of defect identification algorithms to adapt to differences in hardware performance, changes in hardware operating status, and hardware aging and degradation, significantly improving the accuracy and stability of defect identification. Attached Figure Description

[0045] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0046] Figure 1 This is a diagram illustrating the application environment of the defect identification method in one embodiment;

[0047] Figure 2 This is a flowchart illustrating a defect identification method in one embodiment;

[0048] Figure 3 This is a flowchart illustrating the process of obtaining defect identification results in one embodiment;

[0049] Figure 4 This is a flowchart illustrating the process of determining the capability assessment results in one embodiment;

[0050] Figure 5 This is a flowchart illustrating the process of determining a defect identification algorithm in one embodiment;

[0051] Figure 6 This is a flowchart illustrating the optimization of the data acquisition device and defect identification algorithm in one embodiment;

[0052] Figure 7 This is a flowchart illustrating a defect identification method in a preferred embodiment;

[0053] Figure 8 This is a structural block diagram of a defect identification device in one embodiment;

[0054] Figure 9 This is a structural block diagram of a defect identification system in one embodiment. Detailed Implementation

[0055] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0056] The technical background of this application will be explained below:

[0057] Defect identification technology is widely used in various fields such as industrial manufacturing, electronic component inspection, and building quality assessment. Its core objective is to capture the feature information of the target object through various inspection equipment, and then use a preset defect identification algorithm to determine whether the target object has defects and, if so, the type and level of those defects. With the continuous improvement of industrial automation, the requirements for the accuracy and real-time performance of defect identification are becoming increasingly stringent.

[0058] In related technologies, defect identification methods are mostly based on fixed algorithm models and hardware configurations, that is, preset detection algorithm parameters and matching them with fixed models of hardware devices used to capture the feature information of target objects. However, in practical applications, due to various factors, these technologies suffer from low accuracy in defect identification, specifically in the following aspects: 1. The performance of hardware devices varies greatly in different application scenarios. For example, in industrial production lines, different batches of cameras have inherent differences in resolution, frame rate, sensor sensitivity, and sampling frequency. Fixed algorithm models cannot adapt to fluctuations in hardware performance, resulting in unstable quality of the collected raw data, which in turn affects the accuracy of subsequent defect identification; 2. Target objects have diverse shapes, materials, and defect types. When the hardware cannot efficiently capture the key features of a certain type of defect, it is easy to miss or falsely detect defects; 3. Hardware devices may experience aging and wear during long-term operation, leading to a decrease in the accuracy of the collected data, which in turn affects the accuracy of subsequent defect identification.

[0059] In summary, the above description illustrates the main problem in the related technologies: the accuracy of defect identification in the related technologies is low.

[0060] Based on this, this application provides a defect identification method: determining the hardware parameters of a data acquisition device, obtaining a capability assessment result of the acquisition device based on the hardware parameters; determining a defect identification algorithm matching the data acquisition device based on a preset matching relationship and the capability assessment result; processing data acquisition data of a preset object to be identified by the data acquisition device to obtain data to be identified; performing defect identification processing on the data to be identified based on the defect identification algorithm to obtain a defect identification result. This application can improve the accuracy of defect identification. See the following embodiments for details:

[0061] The defect identification method provided in this application embodiment can be applied to, for example... Figure 1In the application environment shown, the data acquisition device 102 communicates with the server 104 via a network. The data storage system can store the data that the server 104 needs to process. The data storage system can be integrated onto the server 104, or it can be placed in the cloud or on another network server. First, the hardware parameters of the data acquisition device are determined. Based on the hardware parameters, the capability assessment results of the data acquisition device are obtained. Based on the preset matching relationship and the capability assessment results, a defect identification algorithm matching the data acquisition device is determined. The data acquisition device performs data acquisition and processing on a preset object to be identified, obtaining the data to be identified. Based on the defect identification algorithm, defect identification processing is performed on the data to be identified to obtain the defect identification result. The data acquisition device 102 can be, but is not limited to, a camera, sensor, etc. The server 104 can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud computing services.

[0062] In one exemplary embodiment, such as Figure 2 As shown, a defect identification method is provided, which can be applied to... Figure 1 Taking server 104 as an example, the explanation includes:

[0063] Step S210: Determine the hardware parameters of the data acquisition device, and obtain the capability assessment result of the acquisition device based on the hardware parameters.

[0064] The data acquisition equipment refers to the hardware devices used to collect data on the object to be identified, such as cameras (which capture the appearance of the object) and sensors (which collect specific signals from the object). The aforementioned hardware parameters include, but are not limited to, basic parameters, real-time operating status parameters, and performance parameters. Basic parameters include hardware model, nominal resolution, nominal sampling frequency, nominal processing speed, and storage capacity. Real-time operating status parameters include the hardware's current operating temperature, power consumption, runtime, and fault warning information. Performance parameters include actual acquisition resolution, actual sampling frequency, data transmission rate, and actual processor processing efficiency. The capability assessment result is a comprehensive evaluation of the data acquisition equipment.

[0065] In this embodiment, the current hardware parameters of the data acquisition device are first obtained, and the quantitative evaluation result of the data acquisition device's capabilities is calculated based on the current hardware parameters and a preset operation logic.

[0066] Step S220: Based on the preset matching relationship and the capability assessment results, determine the defect identification algorithm that matches the data acquisition device.

[0067] The matching relationship is the matching relationship between the defect identification algorithm and the capability assessment results of the data acquisition equipment. That is, each defect identification algorithm is associated with a suitable hardware capability range, which is composed of the capability assessment values ​​of the data acquisition equipment. One range corresponds to one rating.

[0068] In this embodiment, the calculated capability assessment result is compared with each hardware capability range in the preset matching relationship. If the capability assessment result falls within the hardware capability range, the comparison is successful, and the defect identification algorithm corresponding to that hardware capability range is the algorithm adapted to the data acquisition device. The defect identification algorithm includes multiple algorithms, such as feature extraction algorithms, defect classification algorithms, and data preprocessing algorithms. That is, the above-mentioned feature extraction algorithms, defect classification algorithms, and data preprocessing algorithms constitute a defect identification algorithm group. In one embodiment, a set of defect identification algorithms can be associated with a suitable hardware capability range, and correspondingly, a set of suitable defect identification algorithms is determined based on the capability assessment result. In another embodiment, each algorithm can also be associated with a suitable hardware capability range, and correspondingly, the suitable feature extraction algorithm, defect classification algorithm, and data preprocessing algorithm are determined based on the capability assessment result.

[0069] In summary, the appropriate defect identification algorithm can be determined based on the current capability assessment results of the data acquisition equipment.

[0070] Step S230: The data acquisition device performs data acquisition and processing on the preset object to be identified to obtain the data to be identified. Based on the defect identification algorithm, the data to be identified is processed for defect identification to obtain the defect identification result.

[0071] The object to be identified may include a single object or a large number of objects to be identified.

[0072] In this embodiment of the application, the data acquisition device is used to acquire and process the data of the object to be identified, such as video, image, sensor data, etc., and the defect identification data is processed based on the defect identification algorithm selected above to obtain the defect identification result.

[0073] In this embodiment, by collecting multi-dimensional hardware parameters and evaluating hardware capabilities, the defect identification algorithm is accurately matched with the hardware capabilities. The defect identification algorithm is then selected to adapt to different hardware performance differences, changes in hardware operating status, and hardware aging and degradation, which significantly improves the accuracy and stability of defect identification.

[0074] In one exemplary embodiment, such as Figure 3As shown, the data acquisition device collects and processes data from a preset object to be identified, obtaining data to be identified. Based on a defect identification algorithm, the data to be identified is then processed for defect identification, yielding defect identification results, including:

[0075] Step S310: The working parameters of the data acquisition device are calibrated based on the defect identification algorithm, and the adjustable parameters in the defect identification algorithm are adjusted according to the hardware parameters of the data acquisition device.

[0076] Among them, the operating parameters are settings that can be controlled and adjusted during the operation of the data acquisition device and directly affect its data acquisition quality, such as exposure time, gain, aperture, white balance, etc.; the above hardware parameters are inherent attributes of the data acquisition device that cannot be adjusted independently and characterize its basic performance and current status, such as resolution, sensor size, operating time, signal-to-noise ratio, etc.

[0077] In this embodiment, the operating parameters of the data acquisition device can be calibrated based on a defect identification algorithm. This includes correcting parameters such as exposure time, focal length, and frame rate, adjusting sensor sampling frequency and sensitivity, and adjusting the processor's computational resource allocation. Adjusting adjustable parameters in the defect identification algorithm based on the hardware parameters of the data acquisition device includes adjusting feature point thresholds, convolution kernel size, classification thresholds, and regularization parameters. For example, if a high temperature is detected in the camera (i.e., the data acquisition device) sensor, and given that high temperatures increase image noise, the defect identification algorithm can automatically increase the size of the convolution kernel in the filtering algorithm to enhance noise reduction.

[0078] For example, if the data acquisition device is an industrial camera, the exposure time of the industrial camera can be adjusted to 0.02s and the focal length to 50mm according to the defect recognition algorithm. The sampling frequency of the pressure sensor can be adjusted to 1000Hz, and 60% of the processor's computing resources can be allocated to the defect recognition task. Adjustable parameters in the defect recognition algorithm can be adjusted according to the hardware parameters. For example, the feature point threshold of the SIFT (Scale-Invariant Feature Transform) feature extraction algorithm can be adjusted to 0.03 and the convolution kernel size can be adjusted to 3×3. The classification threshold of the CNN (Convolutional Neural Network) defect classification algorithm can be adjusted to 0.85 and the regularization parameter can be adjusted to 0.01 to complete the collaborative calibration.

[0079] Step S320: The data acquisition device is used to acquire and process the data of the object to be identified to obtain the data to be identified. Based on the adjusted defect identification algorithm, the defect identification data is processed to obtain the defect identification result.

[0080] In this embodiment of the application, the data acquisition device is used to acquire and process the data of the object to be identified after calibration to obtain the data to be identified, and the defect identification algorithm is used to process the defect identification data to obtain the defect identification result.

[0081] The embodiments of this application can improve the accuracy and stability of defect identification. Through pre-collaborative calibration, it is ensured that the data acquisition equipment is corrected based on the defect identification algorithm. It also enables the defect detection algorithm to automatically calibrate and compensate according to the working conditions of the data acquisition equipment, thereby improving environmental adaptability and reducing the frequency of manual intervention and maintenance.

[0082] In one exemplary embodiment, the method further includes:

[0083] After obtaining the defect identification results, the performance status of the data acquisition equipment is tested;

[0084] When the performance status of the data acquisition device is detected to match the preset calibration conditions, new hardware parameters of the data acquisition device are collected, new capability assessment results of the data acquisition device are determined based on the new hardware parameters, and a new defect identification algorithm is determined based on the new capability assessment results.

[0085] The performance status includes, but is not limited to, the cumulative running time and frame rate of the data acquisition device.

[0086] In this embodiment, the data acquisition device is continuously monitored. When the performance status of the data acquisition device matches the calibration conditions, the defect identification algorithm is re-matched. This involves acquiring new hardware parameters of the data acquisition device, determining a new capability assessment result based on the new hardware parameters, and then determining a new matching defect identification algorithm based on the new capability assessment result. The calibration conditions can be set by relevant technical personnel according to actual needs. For example, they can be set to trigger the re-matching of the data acquisition device and the defect identification algorithm when the cumulative runtime of the data acquisition device is detected to be greater than or equal to a preset duration threshold (e.g., 1000h), or when the actual frame rate of the data acquisition device is detected to drop to a preset frame rate threshold (e.g., 12fps). Similarly, the adjustable parameters of the data acquisition device and the defect identification algorithm are recalibrated.

[0087] Through the embodiments of this application, the performance of the data acquisition status is continuously monitored, and the defect identification algorithm is adjusted as needed, thereby reducing the impact on the accuracy of defect identification caused by the performance degradation of the data acquisition equipment due to long-term operation.

[0088] In one exemplary embodiment, such as Figure 4As shown, the hardware parameters of the data acquisition device are determined, and the capability assessment results of the acquisition device are obtained based on the hardware parameters, including:

[0089] Step S410: Collect the initial hardware parameters of the data acquisition device, and normalize the initial hardware parameters to obtain the hardware parameters; the hardware parameters include multiple parameter levels.

[0090] The initial hardware parameters are the basic parameters, real-time operating status parameters, and performance parameters of the original data acquisition device. In other words, the parameter hierarchy includes the basic parameter hierarchy, the real-time operating status parameter hierarchy, and the performance parameter hierarchy.

[0091] In this embodiment of the application, the initial hardware parameters of the data acquisition device are collected and normalized to eliminate dimensional differences.

[0092] Step S420: Determine the weights corresponding to each parameter level, and calculate the initial capability assessment results based on the hardware parameters and their corresponding weights.

[0093] The weights correspond to the parameter levels and can be set by relevant technical personnel according to actual needs. Alternatively, the weights corresponding to each parameter level can be determined by the analytic hierarchy process (AHP). For example, the weight of the basic parameter level can be set to 0.3, the weight of the real-time running status parameter level can be set to 0.5, and the weight of the performance parameter level can be set to 0.5.

[0094] The initial capability assessment result of the data acquisition device is obtained by weighted summation based on the various hardware parameters and their corresponding weights.

[0095] For example, the Analytic Hierarchy Process (AHP) is used to determine the weights of basic parameters (0.3), real-time operating status parameters (0.2), and performance parameters (0.5). Each parameter is normalized; for instance, the actual frame rate of the industrial camera (13.8 fps) is normalized to 0.986 (based on a nominal frame rate of 14 fps), and the actual processing efficiency of the processor (92%) is normalized to 0.92. The weighted summation method is used to calculate the overall hardware capability score, which is 0.93. Therefore, the initial capability assessment result is 0.93.

[0096] Step S430: Obtain the stability status and key indicator compliance status of the data acquisition equipment. Based on the initial capability assessment results, stability status, and key indicator compliance status, obtain the capability assessment results.

[0097] Among them, stability refers to the degree of fluctuation of the key status indicators and performance parameters of the data acquisition equipment over a period of time. The smaller the fluctuation, the higher the stability. The compliance status of the above key indicators can be the standard set by relevant technical personnel based on the current specific defect identification task to determine whether one or more key indicators in the data acquisition equipment meet the task requirements. This is generally a binary judgment (compliant / non-compliant).

[0098] In this embodiment, after calculating the initial capability assessment result, the stability and key indicator compliance of the data acquisition device are further obtained. In one embodiment, the initial capability assessment result can be adjusted based on the stability and key indicator compliance. For example, if the stability is average, the stability adjustment coefficient can be determined to be 0.9; if the key indicator meets the requirements, the key indicator adjustment coefficient can be 1. The final capability assessment result can then be a quantitative result, i.e., 0.9 × initial capability assessment result 0.93 × 1. In another embodiment, the initial capability assessment result, stability, and key indicator compliance can be output comprehensively. For example, if the initial capability assessment result of 0.93 belongs to the preset level 2 (preset initial capability assessment result ≥ 0.9 is level 1 performance, 0.8-0.9 is level 2 performance, and initial capability assessment result < 0.8 is level 3 performance), and the data acquisition device is currently operating stably and meets the accuracy requirements of the target detection task, the capability assessment result is obtained. In this case, the capability assessment result is not a quantitative value.

[0099] Through the embodiments of this application, a comprehensive capability assessment result is obtained by integrating the comprehensive capability assessment values, stability status, and key indicator compliance status, thereby enabling the acquisition of a more suitable defect identification algorithm.

[0100] In one exemplary embodiment, such as Figure 5 As shown, based on the preset matching relationship and the capability assessment results, a defect identification algorithm matching the data acquisition equipment is determined, including:

[0101] Step S510: Obtain the matching relationship between multiple preset defect identification algorithms and hardware capability range.

[0102] The preset algorithm library stores multiple preset algorithms, such as SIFT feature extraction algorithm, SURF (Speeded-Up Robust Features) feature extraction algorithm, CNN defect classification algorithm, SVM (Support Vector Machine) defect classification algorithm, Gaussian filtering preprocessing algorithm, and median filtering preprocessing algorithm. Each algorithm is associated with a suitable hardware capability range. In one embodiment, each algorithm is associated with a suitable hardware capability range. For example, the SIFT feature extraction algorithm is suitable for hardware with first-level and second-level capability ranges, the SURF feature extraction algorithm is suitable for hardware with second-level and third-level capability ranges, the CNN defect classification algorithm is suitable for hardware with first-level capability ranges, the SVM defect classification algorithm is suitable for hardware with second-level and third-level capability ranges, and the Gaussian filtering preprocessing algorithm is suitable for hardware with all levels of performance and has a higher computational efficiency than the median filtering preprocessing algorithm. By comparing the hardware capability evaluation results with the algorithm's suitable range, the algorithm combination obtained is SIFT feature extraction algorithm + CNN defect classification algorithm + Gaussian filtering preprocessing algorithm. In another embodiment, a set of defect identification algorithms (including but not limited to feature extraction algorithms, defect classification algorithms, and data preprocessing algorithms) may be associated with a suitable range of hardware capabilities.

[0103] Step S520: Determine the target hardware capability range from multiple hardware capability ranges based on the capability assessment results.

[0104] In this embodiment of the application, the target hardware capability range to which the capability assessment result belongs is determined from multiple hardware capability ranges based on the capability assessment result.

[0105] Step S530: Determine the corresponding defect identification algorithm based on the target hardware capability range.

[0106] In this embodiment of the application, the corresponding defect identification algorithm is determined according to the target hardware capability range. If there are multiple algorithms with the same compatibility, the algorithm with higher computational efficiency is selected first.

[0107] In one exemplary embodiment, such as Figure 6 As shown, the method also includes:

[0108] Step S610: Determine the confidence level corresponding to each defect identification result, and count the target confidence levels among all confidence levels that are greater than the preset confidence threshold.

[0109] Each defect identification result output by the defect identification algorithm corresponds to a confidence level. This confidence level represents the probability assessment value of the correctness of the algorithm's output of a single identification result, and is between 0 and 1. For example, the confidence level of 0.98 for the identification result "there is a metallic foreign object" means that the algorithm is 98% confident that the judgment is correct.

[0110] In this embodiment of the application, the confidence levels corresponding to all defect identification results are statistically analyzed, and the target confidence level that is greater than the preset confidence threshold is determined. The confidence threshold can be set by relevant technical personnel according to actual needs, such as 99%, and the confidence level greater than 99% (i.e. 0.99) is determined as the target confidence level.

[0111] Step S620: When the proportion of the target confidence level in all confidence levels does not reach the preset standard threshold, at least one of the following adjustment schemes is executed: adjusting the defect identification algorithm matched with the data acquisition device, calibrating the working parameters of the data acquisition device, and adjusting the adjustable parameters of the defect identification algorithm.

[0112] In this embodiment of the application, the proportion of the target confidence level in all confidence levels is statistically analyzed. For example, if there are 100 defect identification results, there are 100 corresponding confidence levels. It is assumed that the number of target confidence levels with values ​​greater than the preset confidence level threshold is 96, that is, the proportion of the target confidence level is 96 / 100 (i.e., 0.96).

[0113] The aforementioned standard threshold can be preset by relevant technical personnel, such as being set to 0.95.

[0114] If the target confidence percentage is detected to be greater than or equal to 0.95, then this batch of identification results will be output as the final output. If the target confidence percentage is detected to be less than 0.95, then at least one of the following adjustment schemes needs to be initiated: 1. Re-match the defect identification algorithm corresponding to the data acquisition device; 2. Recalibrate the operating parameters of the data acquisition device; 3. Readjust the adjustable parameters in the defect identification algorithm.

[0115] Step S630: Based on the adjusted data acquisition equipment and defect identification algorithm, re-detect the object to be identified, repeat the above steps, until the proportion of the target confidence score in all confidence scores reaches the standard threshold.

[0116] In this embodiment of the application, when the proportion of the target confidence level in all confidence levels does not reach the preset standard threshold, the data acquisition device and defect identification algorithm are adjusted by the above adjustment method, and the object to be identified is re-detected based on the new data acquisition device and defect identification algorithm. The above steps are repeated until the proportion of the target confidence level in all confidence levels reaches the standard threshold.

[0117] The embodiments of this application can independently detect the identification results. Once a decline in the target confidence ratio is detected, adjustments to the data acquisition equipment and defect identification algorithm can be triggered in a timely manner, ensuring the accuracy of defect identification.

[0118] This application also provides a preferred embodiment of a defect identification method. For example... Figure 7 This is a flowchart illustrating a defect identification method in a preferred embodiment.

[0119] Step S710: Obtain the hardware parameters of the data acquisition device. The hardware parameters include, but are not limited to, basic parameters, real-time operating status parameters, and performance parameters. The basic parameters include the hardware model, nominal resolution, nominal sampling frequency, nominal processing speed, and storage capacity. The real-time operating status parameters include the current operating temperature, power consumption, running time, and fault warning information of the hardware. The performance parameters include the actual acquisition resolution, actual sampling frequency, data transmission rate, and actual processor processing efficiency.

[0120] Step S720: Generate a capability assessment result corresponding to the data acquisition device based on the hardware parameters. The capability assessment result can be presented in the form of quantified numbers, or the quantified numbers can be converted into corresponding levels based on a preset correspondence. Alternatively, the level can be combined with the stability assessment result of the data acquisition device and whether the key indicators meet the standards, and used together as the capability assessment result of the data acquisition device.

[0121] Step S730: Determine the appropriate defect identification algorithm combination based on the capability assessment results. The defect identification algorithm combination includes, but is not limited to, feature extraction algorithms, defect classification algorithms, and data preprocessing algorithms. Each i algorithm is associated with an appropriate capability range / level. Thus, a suitable defect identification algorithm can be determined based on the capability assessment results calculated above.

[0122] Step S740: The working parameters of the data acquisition device are calibrated according to the defect identification algorithm, and the adjustable parameters in the defect identification algorithm are adjusted according to the hardware parameters of the data acquisition device.

[0123] Step S750: The data acquisition device is adjusted to collect and process data of the object to be identified to obtain the data to be identified. Based on the defect identification algorithm, the data to be identified is processed to identify defects to obtain the defect identification result. If the proportion of the target confidence in the defect identification result does not reach the preset standard threshold, the data acquisition device and / or the defect identification algorithm are adjusted. Based on the adjusted data acquisition device and / or the defect identification algorithm, the object to be identified is reprocessed to obtain a new defect identification result. The above steps are repeated until the proportion of the target confidence corresponding to the new identification result reaches the preset standard threshold.

[0124] In step S760, when an abnormal state or performance degradation of the data acquisition device is detected, the process jumps to step S710 to ensure the accuracy of defect identification.

[0125] This application embodiment achieves precise matching between defect identification algorithms and hardware capabilities by collecting multi-dimensional hardware parameters and evaluating hardware capabilities. At the same time, through collaborative calibration and iterative optimization of hardware and algorithms, it effectively adapts to problems such as differences in hardware performance, changes in hardware operating status, and hardware aging and degradation, significantly improving the accuracy and stability of defect identification. In addition, the preset algorithm library supports updates and expansions, improving the versatility and adaptability of the method, and can be widely applied to various defect identification scenarios.

[0126] It should be understood that although the steps in the flowcharts of the above embodiments are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the above embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0127] Based on the same inventive concept, this application also provides a defect identification device for implementing the defect identification method described above. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations in one or more defect identification device embodiments provided below can be found in the limitations of the defect identification method described above, and will not be repeated here.

[0128] In one exemplary embodiment, such as Figure 8 As shown, a defect identification device is provided, comprising:

[0129] Module 81 is used to determine the hardware parameters of the data acquisition device and obtain the capability assessment result of the acquisition device based on the hardware parameters.

[0130] The matching module 82 is used to determine the defect identification algorithm that matches the data acquisition device based on the capability assessment results and a preset matching relationship.

[0131] The generation module 83 is used to collect and process data of the preset object to be identified by the data acquisition device to obtain the data to be identified, and to perform defect identification processing on the data to be identified based on the defect identification algorithm to obtain the defect identification result.

[0132] Each module in the aforementioned defect identification device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0133] In one exemplary embodiment, such as Figure 9 As shown, a defect identification system is provided:

[0134] Hardware acquisition module 91 is used to determine the hardware parameters of the data acquisition device;

[0135] Capability assessment module 92 is used to obtain capability assessment results of the acquisition device based on hardware parameters;

[0136] The algorithm matching module 93 is used to determine the defect identification algorithm that matches the data acquisition device based on the capability assessment results and a preset matching relationship.

[0137] The defect identification module 94 is used to collect and process data of a preset object to be identified by a data acquisition device to obtain the data to be identified, and to perform defect identification processing on the data to be identified based on the defect identification algorithm to obtain the defect identification result.

[0138] In this embodiment, the defect identification system is an integrated system that combines hardware, algorithms, and control logic. Its core objective is to improve the accuracy and stability of defect identification through dynamic adaptive matching of hardware and algorithms.

[0139] The data acquisition module is the fundamental execution unit of the system, comprising an image acquisition device (for capturing visual information of the target object), sensors (for acquiring physical characteristic parameters of the target object, such as pressure and temperature), a processor (for carrying out the computational tasks of each module), and a storage device (for storing algorithm libraries, hardware parameters, recognition data, etc.). The data acquisition module includes a parameter acquisition unit and a data transmission unit. The parameter acquisition unit establishes communication with the aforementioned data acquisition device through a hardware interface, acquiring the hardware parameters, real-time operating status parameters, and performance parameters of the data acquisition device. The data transmission unit transmits the acquired data using a real-time data acquisition protocol.

[0140] The hardware acquisition module is responsible for collecting the hardware parameters of the core hardware at the hardware layer. The capability assessment module constructs an assessment model based on the collected hardware parameters to quantify the current performance and status of the hardware. The algorithm matching module selects algorithm combinations from a pre-set algorithm library that are suitable for the current hardware capabilities. The defect identification module uses data acquisition equipment and algorithms to complete defect detection and classification of the target object.

[0141] In one embodiment, the system further includes a collaborative calibration module and a result optimization module;

[0142] The collaborative calibration module is used to calibrate the working parameters of the data acquisition equipment based on the defect identification algorithm, and to adjust the adjustable parameters in the defect identification algorithm according to the hardware parameters of the data acquisition equipment.

[0143] The result optimization module is used to collect new hardware parameters of the data acquisition device when the performance status of the data acquisition device is detected to match the preset calibration conditions, determine new capability assessment results of the data acquisition device based on the new hardware parameters, and determine a new matching defect identification algorithm based on the new capability assessment results.

[0144] In this embodiment, the collaborative calibration module achieves collaborative optimization of hardware and algorithm by adjusting hardware parameters and key algorithm parameters; the result optimization module continuously improves the system's recognition performance through feedback iteration of recognition accuracy data.

[0145] In one exemplary embodiment, a computer device is provided, including a memory and a processor, the memory storing a computer program, the processor executing the computer program to implement any of the defect identification methods described above.

[0146] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements any of the defect identification methods described above.

[0147] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements any of the defect identification methods described above.

[0148] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0149] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0150] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A defect identification method, characterized in that, The method includes: Determine the hardware parameters of the data acquisition device, and obtain the capability assessment result of the acquisition device based on the hardware parameters; Based on the preset matching relationship and the capability assessment results, a defect identification algorithm matching the data acquisition device is determined. The data acquisition device performs data acquisition and processing on the preset object to be identified to obtain the data to be identified. Based on the defect identification algorithm, the data to be identified is processed to obtain the defect identification result.

2. The method according to claim 1, characterized in that, The step of collecting and processing data on a preset object to be identified using the data acquisition device to obtain data to be identified, and then performing defect identification processing on the data to be identified based on the defect identification algorithm to obtain a defect identification result includes: The working parameters of the data acquisition device are calibrated based on the defect identification algorithm, and the adjustable parameters in the defect identification algorithm are adjusted according to the hardware parameters of the data acquisition device. The data acquisition device is used to acquire and process the data of the object to be identified to obtain the data to be identified. The data to be identified is then processed for defect identification based on the adjusted defect identification algorithm to obtain the defect identification result.

3. The method according to claim 1, characterized in that, The method further includes: Upon obtaining the defect identification results, the performance status of the data acquisition device is tested. When the performance status of the data acquisition device is detected to match the preset calibration conditions, new hardware parameters of the data acquisition device are collected, a new capability assessment result of the data acquisition device is determined based on the new hardware parameters, and a new defect identification algorithm is determined based on the new capability assessment result.

4. The method according to any one of claims 1 to 3, characterized in that, The process of determining the hardware parameters of the data acquisition device and obtaining a capability assessment result of the acquisition device based on the hardware parameters includes: The initial hardware parameters of the data acquisition device are collected, and the initial hardware parameters are normalized to obtain the hardware parameters; the hardware parameters include multiple parameter levels. Determine the weights corresponding to each parameter level, and calculate the initial capability assessment results based on the hardware parameters and their corresponding weights. The stability and key performance indicators of the data acquisition device are obtained, and the capability assessment result is obtained based on the initial capability assessment result, stability and key performance indicators.

5. The method according to any one of claims 1 to 3, characterized in that, The defect identification algorithm, which determines the matching algorithm with the data acquisition device based on the capability assessment results according to a preset matching relationship, includes: Obtain the matching relationship between multiple preset defect identification algorithms and the hardware capability range; The target hardware capability range is determined from multiple hardware capability ranges based on the capability assessment results; The corresponding defect identification algorithm is determined based on the target hardware capability range.

6. The method according to any one of claims 1 to 3, characterized in that, The method further includes: Determine the confidence level corresponding to each defect identification result, and count the target confidence levels among all confidence levels that are greater than the preset confidence threshold; When the percentage of the target confidence level among all confidence levels does not reach the preset standard threshold, at least one of the following adjustment schemes is executed: adjusting the defect identification algorithm matched with the data acquisition device, calibrating the working parameters of the data acquisition device, and adjusting the adjustable parameters of the defect identification algorithm; Based on the adjusted data acquisition equipment and defect identification algorithm, the object to be identified is re-detected, and the above steps are repeated until the proportion of the target confidence level in all confidence levels reaches the standard threshold.

7. A defect identification device, characterized in that, The device includes: The determination module is used to determine the hardware parameters of the data acquisition device and obtain the capability evaluation result of the acquisition device based on the hardware parameters. The matching module is used to determine a defect identification algorithm that matches the data acquisition device based on the capability assessment results and a preset matching relationship. The generation module is used to collect and process data of a preset object to be identified by the data acquisition device to obtain data to be identified, and to perform defect identification processing on the data to be identified based on the defect identification algorithm to obtain defect identification results.

8. A defect identification system, characterized in that, The system includes a data acquisition device, a hardware information acquisition module, a capability assessment module, an algorithm matching module, and a defect identification module; The hardware acquisition module is used to determine the hardware parameters of the data acquisition device; The capability assessment module is used to obtain the capability assessment result of the acquisition device based on the hardware parameters; The algorithm matching module is used to determine a defect identification algorithm that matches the data acquisition device based on the capability assessment results and a preset matching relationship. The defect identification module is used to collect and process data on a preset object to be identified by the data acquisition device to obtain data to be identified, and to perform defect identification processing on the data to be identified based on the defect identification algorithm to obtain a defect identification result.

9. The system according to claim 8, characterized in that, The system also includes a collaborative calibration module and a result optimization module; The collaborative calibration module is used to calibrate the operating parameters of the data acquisition device based on the defect identification algorithm, and to adjust the adjustable parameters in the defect identification algorithm according to the hardware parameters of the data acquisition device. The result optimization module is used to collect new hardware parameters of the data acquisition device when the performance status of the data acquisition device is detected to match the preset calibration conditions, determine new capability assessment results of the data acquisition device based on the new hardware parameters, and determine a new matching defect identification algorithm based on the new capability assessment results.

10. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

Images