A method, system, medium and program product for visual identification of good products

Abstract

A method, system, medium, and program product for visual recognition of qualified products, relating to the field of image data processing, comprising: determining the inspection type of the product to be traced; upon a process inspection request, obtaining information on the abnormal process that triggers the inspection, calculating the process correlation between each completed process and the abnormal process, and generating a sequence to be inspected; upon a final inspection request, extracting defect feature data of the product to be traced, determining a set of associated processes from a process flow database, and sorting and generating a sequence to be inspected; sequentially obtaining images of the processes to be inspected of the top-ranked processes in the sequence to be inspected, and inputting them into a parallel inspection channel for inspection; when the result of the parallel inspection channel is passed, updating the sequence to be inspected, and obtaining the image of the first process for inspection; and when the result of the parallel inspection channel is abnormal, determining the abnormal process and stopping inspection. Implementing this application can improve the efficiency of problem location based on visual recognition.

Description

Technical Field

[0001] The present application relates to the field of image data processing, and in particular to a method, system, medium and program product for visually identifying good products. Background Art

[0002] The production process for laptop cases involves multiple steps, including material selection, mold design, injection molding, machining, surface treatment, painting, and quality inspection. As demand for laptops continues to grow, production scale continues to expand, raising the bar for efficiency and quality control across each process. In actual production, due to the numerous and interconnected processes, any quality issues that arise require rapid identification and appropriate action to ensure the continued efficient operation of the production line.

[0003] In related technologies, multiple inspection points are set up on the notebook casing production line. Inspectors inspect the products according to pre-set inspection procedures, record relevant information when anomalies are found, and analyze possible causes. At the same time, automated inspection equipment is installed at key processes to monitor parameters such as product appearance and dimensions in real time, and the inspection data is stored in a database for subsequent analysis.

[0004] However, as the scale of production continues to expand, especially when the production line is long, due to the process correlation between each process, problems in a certain process will be transmitted to subsequent processes, and the efficiency of problem location based on visual recognition is low. Summary of the Invention

[0005] The present application provides a method, system, medium and program product for visually identifying good products, which are used to improve the efficiency of problem location based on visual recognition.

[0006] In a first aspect, the present application provides a method for visual recognition of defective products, which is applied to a visual recognition system, and the method includes: in response to a product traceability application, determining the inspection type of the product to be traced; the inspection type includes a process inspection request or a final inspection request; when the inspection type is a process inspection request, obtaining the abnormal process information that triggers the inspection; calculating the process correlation between each completed process and the abnormal process, and sorting each completed process based on the process correlation to generate a sequence to be inspected; when the inspection type is a final inspection request, extracting defect feature data of the product to be traced; based on the defect feature data, determining a set of associated processes from a process flow database, and sorting them to generate a sequence to be inspected; starting multiple parallel inspection channels, and sequentially obtaining the process images to be inspected of each top-ranked process in the sequence to be inspected, and inputting the process images to be inspected into the parallel inspection channel for inspection; when the parallel inspection channel completes the current process inspection and the result is passed, updating the sequence to be inspected, and obtaining the image of the first process in the updated sequence to be inspected for inspection; when the parallel inspection channel detects a process abnormality, determining the corresponding process as an abnormal process, and stopping the inspection of all parallel inspection channels.

[0007] In the above embodiment, the visual recognition system is able to determine the inspection type, optimize the inspection sequence based on process correlation during process inspection, use defect feature data to spot-check related processes during final inspection, and simultaneously process the top-ranked process images by starting multiple parallel inspection channels. It dynamically updates the sequence when the inspection passes and stops the inspection in time when an anomaly is found, which significantly improves the efficiency of problem location and reduces unnecessary waste of inspection resources.

[0008] In combination with some embodiments of the first aspect, in some embodiments, before the step of determining the detection type of the product to be traced in response to a product traceability application, the method also includes: obtaining product monitoring data of multiple monitoring devices in the product production process; performing feature recognition on the product monitoring data, determining marked products whose product feature values ​​deviate from a preset threshold range, and generating abnormal prompt information; generating a product traceability application after a preset waiting time, or when receiving a product traceability instruction entered by the user after viewing the abnormal prompt information.

[0009] In the above embodiment, the visual recognition system obtains product monitoring data from multiple monitoring devices in real time, compares the characteristic values ​​with preset thresholds, promptly detects anomalies and generates prompt information, and automatically triggers the traceability application after a preset waiting time or user confirmation, thereby achieving early warning and rapid response to anomalies.

[0010] In combination with some embodiments of the first aspect, in some embodiments, when the parallel detection channel completes the current process detection and the result is passed, the sequence to be detected is updated, and the image of the first process in the updated sequence to be detected is obtained for detection. The steps specifically include: when the parallel detection channel completes the current process detection and the result is passed, generating detection pass data based on the current process; based on the detection pass data, removing the associated processes in the sequence to be detected to obtain the optimized sequence to be detected; re-sorting the optimized sequence to be inspected, and obtaining the sorted image of the first process to the parallel detection channel for detection.

[0011] In the above embodiment, the visual recognition system generates detection pass data when the detection passes, and dynamically removes the associated processes in the sequence to be detected based on this data, and reorders the optimized sequence. This sequence optimization mechanism based on the detection results avoids repeated detection, improves detection efficiency, and ensures the integrity and accuracy of the detection.

[0012] In combination with some embodiments of the first aspect, in some embodiments, after the step of generating detection pass data based on the current process when the parallel detection channel completes the current process detection and the result is passed, the method also includes: when other parallel detection channels complete the corresponding current process detection and the result is passed, updating the detection pass data based on the corresponding current process detection.

[0013] In the above embodiment, the visual recognition system supports multiple parallel detection channels to update the detection data at the same time, realizes the real-time summary of the detection results, improves the concurrent processing capability of the system, and speeds up the problem location speed.

[0014] In combination with some embodiments of the first aspect, in some embodiments, before the step of removing the associated processes in the sequence to be detected based on the detection pass data to obtain the optimized sequence to be detected, the method also includes: obtaining the confidence value of the detection pass data of the current process; when the confidence value is lower than the preset confidence threshold, obtaining the first process of the sequence to be detected to the parallel detection channel for detection, and adding the current process to the first position of the sequence to be detected.

[0015] In the above embodiment, the visual recognition system evaluates the reliability of the detection results through confidence values, and automatically adjusts the detection order when the confidence is low, ensuring that key suspicious processes are inspected first, thereby improving the accuracy of problem location.

[0016] In combination with some embodiments of the first aspect, in some embodiments, the step of starting multiple parallel detection channels, and sequentially obtaining images of the process to be inspected of the top-ranked processes in the sequence to be inspected, and inputting the images of the process to be inspected into the parallel detection channels for inspection, specifically includes: starting multiple parallel detection channels, and obtaining hardware parameter information corresponding to each parallel detection channel; performing computing power grading on each parallel detection channel according to the hardware parameter information to obtain computing power grading results; sequentially obtaining images of the process to be inspected of the top-ranked processes in the sequence to be inspected, and calculating the estimated detection time of the images of the process to be inspected; and performing channel allocation on the images of the process to be inspected based on the computing power grading results and the estimated detection time.

[0017] In the above embodiment, the visual recognition system classifies the computing power of the detection channels based on hardware parameters, and allocates channels according to the expected detection time of the process images, thereby achieving load balancing of the detection tasks.

[0018] In combination with some embodiments of the first aspect, in some embodiments, when a parallel detection channel detects a process abnormality, the corresponding process is determined as an abnormal process, and after the step of stopping the detection of all parallel detection channels, the method also includes: obtaining the detection image and detection parameters of the abnormal process, and generating an abnormality analysis report; generating abnormal detection results and parameter adjustment suggestions based on the abnormality analysis report, and receiving process adjustment instructions input by the user; adjusting the process parameters of the production equipment of the abnormal process according to the process adjustment instructions.

[0019] In the above embodiment, the visual recognition system automatically generates anomaly analysis reports and parameter adjustment suggestions, and supports parameter adjustment of production equipment through process adjustment instructions, thereby improving the automation level and response speed of the production line.

[0020] In second aspect, an embodiment of the present application provides a visual recognition system, which includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to enable the visual recognition system to execute the method described in the first aspect and any possible implementation method of the first aspect.

[0021] In a third aspect, an embodiment of the present application provides a computer program product comprising instructions, which, when the computer program product is run on a visual recognition system, enables the visual recognition system to execute the method described in the first aspect and any possible implementation of the first aspect.

[0022] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium comprising instructions. When the instructions are executed on a visual recognition system, the visual recognition system executes the method described in the first aspect and any possible implementation of the first aspect.

[0023] It is understandable that the visual recognition system provided in the second aspect, the computer program product provided in the third aspect, and the computer storage medium provided in the fourth aspect are all used to execute the methods provided in the embodiments of the present application. Therefore, the beneficial effects that can be achieved can be referenced to the beneficial effects of the corresponding methods and will not be repeated here.

[0024] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages:

[0025] 1. Due to the adoption of a dual-path detection strategy based on detection type and a combined mechanism of multi-channel parallel detection and dynamic sequence optimization, the system can select the optimal detection path according to different scenarios and improve detection efficiency through parallel processing, effectively solving the problems of single detection path and low serial processing efficiency in existing technologies, thereby achieving efficient utilization of detection resources and rapid problem location; by determining the detection type to distinguish the processing path, optimizing the detection sequence based on process correlation during process detection, using defect feature data to spot-check related processes during final inspection, and starting multiple parallel detection channels to simultaneously process process images, the efficiency of problem location is significantly improved.

[0026] 2. By utilizing real-time feature recognition and an automatic triggering mechanism based on monitoring data, the system can promptly detect anomalies during the production process and automatically initiate the traceability process. This effectively addresses the issues of delayed anomaly detection and time-consuming manual intervention in existing technologies, thereby enabling early warning and rapid response to anomalies. By continuously acquiring monitoring device data and performing feature recognition, products with abnormal characteristic values ​​are identified. When an anomaly is detected, the system automatically generates a prompt message, improving the timeliness of anomaly handling.

[0027] 3. Due to the adoption of a sequence optimization mechanism and confidence assessment strategy based on test results, the system can dynamically adjust the test order and ensure the reliability of the test results, effectively solving the problem of fixed test sequence and repeated test that wastes resources in the existing technology, thereby achieving a dual improvement in test efficiency and accuracy; after passing the test, the test pass data will be generated and the sequence to be tested will be optimized accordingly, unnecessary related processes will be removed, thus avoiding waste of resources and ensuring the reliability of the test. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 This is a flow chart of a method for visually identifying good products in an embodiment of the present application;

[0029] Figure 2 This is another flowchart of the method for visually identifying good products in an embodiment of the present application;

[0030] Figure 3 It is a schematic diagram of the physical device structure of the visual recognition system in the embodiment of the present application. DETAILED DESCRIPTION

[0031] The terms used in the following examples of the present application are only for the purpose of describing specific embodiments and are not intended to limit the present application. As used in the specification of the present application, the singular expressions "a", "an", "above", "the", and "this" are intended to include plural expressions as well, unless the context clearly indicates otherwise. It should also be understood that the term "and / or" used in the present application refers to any or all possible combinations of one or more of the listed items.

[0032] In the following, the terms "first" and "second" are used for descriptive purposes only and should not be understood to imply or suggest relative importance or implicitly indicate the number of the technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the embodiments of this application, unless otherwise specified, "plurality" means two or more.

[0033] For ease of understanding, the application scenarios of the embodiments of the present application are introduced below.

[0034] The production of laptop cases involves multiple steps, including injection molding, machining, surface treatment, and painting. As market demand grows and production scale expands, production lines are becoming longer and the interdependencies between processes are becoming increasingly complex. For example, temperature control issues in the injection molding process can lead to dimensional deviations in the product, impacting subsequent surface treatment. Abnormalities in the painting process may stem from improper surface treatment in a previous step. When product quality issues are discovered, the specific problematic process must be quickly identified and addressed promptly to prevent further losses from continued defective products. This places higher demands on quality monitoring and problem tracing during the production process.

[0035] In related technologies, real-time monitoring of product quality can be achieved by employing fixed inspection points and pre-set inspection procedures. Specifically, multiple manual inspection stations and automated inspection equipment are deployed on the production line, performing individual inspections according to a predetermined inspection sequence and standards. Any anomalies detected are recorded and analyzed. The following describes a scenario in which the related art visual recognition method for good product quality is used.

[0036] In the existing technology, quality control is usually carried out by combining manual inspection with automatic inspection. Multiple inspection points are set up on the production line, and inspectors inspect the products according to preset inspection procedures. At the same time, automatic inspection equipment is installed in key processes such as injection molding and painting to monitor product appearance, size and other parameters in real time. When an abnormality is found, the inspector needs to judge the possible source of the problem based on experience and check the relevant processes one by one. For example, when scratches are found on the surface of the product, it is necessary to check multiple processes such as painting, polishing, and surface treatment in turn. Each process usually contains multiple sub-processes, and the time period involved is relatively long. This serial inspection method is inefficient and prone to omissions, making it difficult to meet the needs of rapid production.

[0037] The method for visually identifying good products in the embodiments of this application, through dynamic process correlation analysis and a multi-channel parallel detection mechanism, achieves rapid location of anomalies and improves detection efficiency. The following describes a scenario in which the method for visually identifying good products in this application is used.

[0038] After adopting the visual recognition method of the present application, the system can automatically analyze the process correlation based on the specific characteristics of the anomaly and determine the process sequence that is most likely to cause problems. For example, when a scratch is detected on the surface of a product, the system will analyze the morphological characteristics of the scratch and, combined with the empirical data in the process database, quickly determine that it may be related to the surface treatment and painting processes. Then, multiple parallel detection channels are started to detect these processes at the same time. If an anomaly is found in a certain detection channel, the system will immediately determine the problematic process; if the test passes, the sequence to be inspected will be dynamically updated. This intelligent parallel detection method significantly improves the efficiency of problem location.

[0039] It can be seen that the visual recognition method of good products in the embodiment of the present application can not only realize automatic monitoring of product quality, but also effectively solve the problem of low efficiency of traditional serial detection, thereby realizing intelligent quality control of the production process.

[0040] For ease of understanding, the following describes the process of the method provided by this implementation in combination with the above scenario. Figure 1 , which is a flow chart of the method for visually identifying good products in an embodiment of the present application.

[0041] S101. In response to a product traceability application, determine a test type for the product to be traced.

[0042] A product traceability request refers to a traceability test request triggered manually by the monitoring system or the user during the production process. The "product to be traced" represents the specific product instance requiring quality traceability and can be uniquely identified by information such as the product number and production batch. The test type indicates the specific scenario for the current traceability test and includes two types: in-process test requests and final inspection requests. In-process test requests trigger real-time testing when an anomaly is discovered during the production process. Final inspection requests are the final quality inspection performed after the product has completed all production steps.

[0043] After receiving a product traceability application, the visual recognition system must first determine the detection type in order to select an appropriate detection strategy. Specifically, the visual recognition system will determine the current detection type based on information such as the trigger source, current product status, and production stage carried in the traceability application. If the traceability application is triggered by abnormal monitoring during the production process, and the product has not yet completed all production processes, it is determined to be a process detection request; if the traceability application is from the quality inspection link after the product is completed, or the product has completed all production processes, it is determined to be a final inspection request. The visual recognition system will also verify the legitimacy of the traceability application, check the integrity of the necessary parameters, and record the timestamp and basis information for determining the detection type.

[0044] In some embodiments, the determination of the detection type can be achieved in a variety of ways: Optionally, the visual recognition system can establish a decision tree model, take the various characteristic parameters of the traceability application as input, and output the detection type after multi-layer conditional judgment, specifically including: extracting characteristic parameters such as trigger time, product status, and number of completed processes in the traceability application; inputting these parameters into a pre-trained decision tree model; and determining the detection type based on the output of the decision tree. Optionally, the visual recognition system can use a rule engine to determine the detection type, specifically including: pre-setting a series of judgment rules and weights; performing feature extraction and rule matching on the traceability application; and calculating the final detection type based on the matching results and rule weights. It is understandable that other machine learning methods or expert systems can also be used to determine the detection type, which is not limited here.

[0045] In the process of determining the inspection type, there may be situations where the traceability application information is incomplete or the parameter values ​​are abnormal. To address this, the visual recognition system uses multiple verification and fault-tolerance mechanisms: first, it verifies whether each parameter is within a reasonable range through preset parameter validity check rules; for missing non-critical parameters, it uses default values ​​or historical data to supplement them; for abnormal parameter values, it uses contextual information and historical experience to make corrections; when it is impossible to accurately determine the inspection type, it gives priority to a more stringent inspection strategy, that is, it classifies it as a final inspection request. For example, when product status information is missing, the visual recognition system will query the production record of the product. If it finds that the process completion rate of the most recent record exceeds 95%, it will be determined as a final inspection request.

[0046] S102: When the detection type is a process detection request, obtain abnormal process information that triggers the detection.

[0047] A process detection request is a detection type triggered when an anomaly is discovered during the production process. Abnormal process information refers to process data detected during the production process that does not meet quality standards, including the process ID, anomaly type, abnormal parameter values, and occurrence time. A triggered detection is a detection action initiated automatically by the anomaly monitoring system or manually by an operator. Process information includes complete data such as basic process attributes, quality parameters, equipment status, and operation records.

[0048] After determining that the current request is a process inspection, the visual recognition system needs to obtain information about the specific abnormal process that triggered the inspection. Specifically, the visual recognition system extracts a complete record of the abnormal process from the abnormality monitoring system's database, including the specific process link where the abnormality occurred, the specific manifestation of the abnormality, the degree of deviation of the abnormal parameters, the time point when the abnormality occurred, the duration of the abnormality, relevant equipment status data, operator records, and other information. The visual recognition system also collects operational data from related processes before and after the abnormal process to establish a complete abnormality context information set.

[0049] In some embodiments, the acquisition of abnormal process information can be achieved through a variety of methods: Optionally, the visual recognition system can monitor the quality parameters of each process on the production line in real time, and automatically record and upload abnormal information when the parameters deviate from the preset threshold, including setting multi-level warning thresholds, real-time acquisition of parameter data, calculation of parameter deviation, determination of abnormality level, and recording of abnormal data; Optionally, the visual recognition system can establish an abnormal feature library and combine it with a machine learning algorithm to identify abnormal patterns, including collecting historical abnormal data, extracting abnormal features, training recognition models, matching abnormal patterns in real time, and outputting abnormal information. It is understandable that other data collection and analysis methods can also be used to achieve the acquisition and processing of abnormal process information, which are not limited here.

[0050] When acquiring information about abnormal processes, incomplete data collection or distorted abnormal data can be problematic. To address this, the visual recognition system employs a data completion and verification mechanism: Multiple data collection channels are set up to ensure data collection reliability; redundant data storage is used to prevent data loss; data consistency verification rules are established to identify and process abnormal data; and when data is incomplete, data from similar operating conditions is interpolated and supplemented. For example, if a sensor collecting a quality parameter fails, the system automatically switches to a backup sensor and corrects the abnormal data collected based on historical data.

[0051] S103: Calculate the process correlation between each completed process and the abnormal process, and sort each completed process based on the process correlation to generate a sequence to be detected.

[0052] Process relevance represents the degree of mutual influence between different processes within the process flow. It includes quantitative indicators such as direct inter-process correlation, correlation between process parameters, and the transfer relationship between quality characteristics. Completed processes refer to all production processes completed before the test is triggered. Sorting prioritizes processes based on the calculated process relevance. The queue for inspection represents the process queue sorted by inspection priority.

[0053] After obtaining information about abnormal processes, the visual recognition system needs to calculate the degree of correlation between each process and the abnormal process in order to determine the source process that is most likely to cause the current abnormality. Specifically, the visual recognition system first extracts basic data such as the topological relationship between processes, the process parameter influence matrix, and the quality feature transfer chain from the process flow database. Then, combined with the specific characteristics of the abnormal process, the influence weight of each completed process on the abnormal process is calculated. The calculation process takes into account multiple dimensions such as direct influence, indirect influence, and cumulative influence between processes, and introduces a time series attenuation factor to adjust the degree of influence of processes completed in different time periods. Finally, the processes are sorted based on the calculated process correlation to generate a sequence to be detected.

[0054] In some embodiments, the calculation and ranking of process correlation can be achieved in a variety of ways: Optionally, the visual recognition system can establish a process correlation network model and calculate the correlation strength between processes through a graph theory algorithm, specifically including: constructing a process correlation graph, setting edge weights, calculating the shortest path, evaluating influence transfer, and summarizing correlation strength; Optionally, the visual recognition system can adopt a multi-factor weighted scoring method, comprehensively considering factors such as process parameter correlation, quality feature consistency, and historical abnormality correlation, specifically including: determining scoring factors, setting weight coefficients, calculating individual scores, weighted summation, and standardization. It is understandable that other mathematical models or algorithms can also be used to achieve quantitative calculation of process correlation, which is not limited here.

[0055] The process of calculating process relevance often presents the problem of high computational complexity due to the large number of processes. To address this, the visual recognition system employs a hierarchical computation strategy: First, a coarse-grained classification of processes is performed to identify those with the highest correlation with abnormal features. Then, fine-grained relevance calculations are performed within these relevant process categories. Finally, the results are combined and normalized. The system also employs parallel computing and caching mechanisms to improve computational efficiency. For example, for common abnormality types, the system pre-calculates and stores process relevance templates, requiring only local updates and adjustments during actual calculations.

[0056] S104: When the inspection type is a final inspection request, extract defect feature data of the product to be traced.

[0057] The final inspection request refers to the final quality inspection process after a product completes all production steps. Defect characteristic data refers to a quantitative description of various quality defects discovered during the final inspection process, including multi-dimensional characteristic parameters such as defect type, location, size, and severity. The extraction process involves obtaining a complete set of defect characteristic information through image processing, data analysis, and other means. Products to be traced refer to specific individual products requiring quality traceability analysis.

[0058] After confirming a request for final inspection, the visual recognition system needs to comprehensively collect and analyze the product's defect characteristics. Specifically, the system first uses the image acquisition equipment at the final inspection station to obtain high-definition, multi-angle images of the product. It then uses image processing algorithms to extract surface defect characteristics, including visible defects such as color difference, scratches, bumps, and bubbles. This system also combines data from other inspection methods, such as dimensional inspection and performance testing, to form a complete defect feature dataset. For each detected defect, the system records its specific parameters and compares them with product quality standards to assess the severity of the defect.

[0059] In some embodiments, defect feature data extraction can be achieved through a variety of methods: Optionally, the visual recognition system can use a deep learning model for defect detection and feature extraction, specifically including: preprocessing image data, using a pretrained neural network model for defect recognition, extracting morphological features of the defect area, calculating defect parameters, and generating feature vectors; Optionally, the visual recognition system can construct a feature extraction process based on traditional image processing methods, specifically including: image enhancement, edge detection, region segmentation, feature calculation, and parameter statistics. It is understood that other computer vision technologies or signal processing methods can also be used to achieve defect feature extraction and quantification, which are not limited here.

[0060] During defect feature extraction, inconsistent feature representations of different defect types can be encountered. To address this, the visual recognition system employs a standardized feature representation framework: establishing a unified defect feature description system to map defect features from different sources and types into a standard feature space; designing feature conversion rules to ensure effective integration of data from different inspection devices; and constructing a feature similarity measurement method to support comparative analysis across defect types. For example, for products with both dimensional deviations and cosmetic defects, the system uses feature conversion to uniformly represent each defect type as a standardized feature vector, facilitating subsequent correlation analysis.

[0061] S105 . Based on the defect feature data, determine a set of related processes from a process flow database, and sort them to generate a sequence to be inspected.

[0062] The process flow database is a data system that stores complete product production process information, including process flow, process parameters, quality characteristics, equipment parameters, and other data. The associated process set represents a combination of production processes that have a causal or correlational relationship with the current defect characteristics. Process sorting prioritizes processes based on the strength of their correlation. The pending inspection sequence represents the prioritized queue of processes awaiting traceability analysis.

[0063] After acquiring complete defect signature data, the visual recognition system needs to identify key processes that may have caused the defect from historical production data. Specifically, the system first matches the defect signature data with defect patterns in the process knowledge base to identify potential causal processes. It then analyzes the parameter transfer relationships within the process flow to determine the mechanisms by which these processes influence the current defect. It then calculates the strength of the association between each relevant process and the defect, taking into account both direct and indirect influences. Finally, it ranks the processes based on the strength of the association, forming a prioritized sequence for inspection.

[0064] In some embodiments, process association analysis and ranking can be achieved through a variety of methods: Optionally, the visual recognition system can establish a defect-process association model and analyze the association rules in historical data through data mining methods, specifically including: constructing a feature-process association matrix, calculating support and confidence, extracting strong association rules, evaluating rule reliability, and generating a set of associated processes; Optionally, the visual recognition system can use causal reasoning methods to construct a causal network based on process expert knowledge, specifically including: establishing a process causal diagram, setting conditional probabilities, performing probabilistic reasoning, calculating posterior probabilities, and identifying key processes. It is understandable that other intelligent analysis methods can also be used to achieve the discovery and quantification of process associations, which are not limited here.

[0065] During process correlation analysis, incorrect correlations can lead to inaccurate detection sequences. To address this, the visual recognition system employs a multi-verification mechanism: first, historical case verification checks whether the analysis results align with known defect-process relationships; then, process expert rules are used to verify that the correlation results adhere to process logic; and finally, cross-validation compares the results obtained by different analysis methods. For example, if the system discovers a high correlation between a process and a specific defect, it automatically retrieves historical data for anomaly records related to that process to verify the universality and reliability of this correlation.

[0066] S106 , starting multiple parallel inspection channels, and sequentially acquiring images of the processes to be inspected that rank top in the sequence to be inspected, and inputting the images of the processes to be inspected into the parallel inspection channels for inspection.

[0067] Parallel inspection channels represent multiple processing units capable of simultaneously executing image inspection tasks, encompassing a combination of hardware resources and software algorithms. Inspected process images refer to image data of products from processes requiring quality inspection. Top rankings refer to the important processes selected based on the priority ranking results determined in previous steps. Image input refers to the process of transmitting process image data to the inspection unit for processing.

[0068] After determining the sequence to be inspected, the visual recognition system needs to efficiently complete the image inspection tasks for multiple processes. Specifically, the system first evaluates available computing resources and initializes multiple parallel inspection channels. It then assigns a corresponding inspection algorithm and parameter configuration to each inspection channel. It then extracts process images from the sequence to be inspected in order of priority and assigns them to the most appropriate inspection channel based on their characteristics and inspection requirements. Simultaneously, it monitors the operating status of each inspection channel and dynamically adjusts the task allocation strategy to ensure maximum inspection efficiency. The system also collects the inspection progress and results of each channel in real time to ensure data synchronization and consistency.

[0069] In some embodiments, the scheduling and execution of parallel detection can be achieved in a variety of ways: Optionally, the visual recognition system can adopt a dynamic load balancing strategy to allocate tasks based on the performance characteristics of the detection channel and the current load situation, specifically including: evaluating image complexity, calculating detection difficulty, estimating processing time, selecting the optimal channel, and dynamically adjusting allocation; Optionally, the visual recognition system can establish a hierarchical detection framework to divide the detection task into subtasks of different granularities for parallel processing, specifically including: task decomposition, resource allocation, parallel execution, result merging, and performance optimization. It is understandable that other parallel computing or task scheduling methods can also be used to achieve efficient image detection, which is not limited here.

[0070] During parallel inspection, uneven performance across inspection channels can lead to reduced processing efficiency. To address this, the visual recognition system employs an adaptive scheduling mechanism: It monitors the processing capacity and efficiency of each inspection channel in real time; establishes a performance evaluation model to predict the processing time for different image types; dynamically adjusts the task allocation strategy based on the evaluation results; and promptly reallocates tasks when performance bottlenecks arise. For example, if the processing speed of a particular inspection channel is significantly lower than that of other channels, the system automatically reduces the number of tasks assigned to that channel and shifts tasks to the higher-performing channel.

[0071] S107 , when the parallel inspection channel completes the current process inspection and the result is passed, the sequence to be inspected is updated, and the image of the first process in the updated sequence to be inspected is obtained for inspection.

[0072] A "pass" result indicates that the product quality inspection for the current process meets the preset standards. Updating the pending inspection sequence dynamically adjusts the queue of pending processes based on the inspection results. The first process image represents the inspection image corresponding to the highest-priority process in the updated sequence. The sequence update process includes deleting processes that have passed inspection, recalculating the priorities of remaining processes, and adjusting the inspection order.

[0073] After receiving a positive test result, the visual recognition system needs to promptly adjust subsequent testing strategies. Specifically, the system first verifies the reliability of the test results, including whether the test parameters are within the valid range, whether the test process is stable, and whether the confidence level of the test results meets the standards. It then analyzes the relationship between the current process and other processes in the sequence to be tested, assessing the impact of the positive test on other processes. It then updates the correlation weights between processes and recalculates the inspection priority of each process. Finally, it generates an updated sequence to be tested and immediately initiates testing of the new top process. The system also saves detailed test results for subsequent quality traceability analysis.

[0074] In some embodiments, dynamic updates of the sequence to be inspected can be achieved in a variety of ways: Optionally, the visual recognition system can establish a process dependency network model and dynamically adjust the inspection priority through a graph structure update algorithm, specifically including: constructing a process dependency graph, calculating node influence, updating edge weights, reordering the sequence, and optimizing the path; Optionally, the visual recognition system can use a heuristic algorithm to adjust the inspection strategy based on inspection history and empirical rules, specifically including: analyzing historical patterns, extracting decision rules, evaluating adjustment plans, executing sequence updates, and verifying the update effects. It is understood that other sequence optimization or scheduling algorithms can also be used to achieve efficient updates of the sequence to be inspected, which are not limited here.

[0075] During the sequence update process, local optimality can lead to decreased detection efficiency. To address this, the visual recognition system employs a global optimization strategy: establishing a multi-objective evaluation system that comprehensively considers factors such as detection efficiency, resource utilization, and result reliability; introducing a random perturbation mechanism to avoid falling into local optimality; and setting dynamic adjustment thresholds to improve processing speed while ensuring detection quality. For example, if the system detects that detection efficiency has not improved after multiple consecutive sequence updates, it automatically triggers a global re-optimization mechanism to re-evaluate the priorities of all pending inspection processes.

[0076] S108. When a parallel detection channel detects a process abnormality, the corresponding process is determined as an abnormal process, and detection of all parallel detection channels is stopped.

[0077] A process abnormality indicates a state where the test results do not meet preset quality standards or process requirements. An abnormal process refers to the specific production link where a quality issue was discovered during testing. Stopping testing interrupts all current testing channels and saves the test status. The process of stopping parallel testing channels includes tasks termination, state saving, and resource release. The process of determining test results includes steps such as abnormality assessment, confidence calculation, and multi-dimensional verification.

[0078] When the visual recognition system finds a process abnormality in any detection channel, it needs to respond quickly and take corresponding measures. This step can be implemented at any time, that is, step S108 can be detected and executed after step S106, that is, after the first round of task allocation of the parallel detection channel, or it can be detected and executed after step S107, that is, after the second round of task allocation of the parallel detection channel; it can be simply deduced that step S107 can be plural, that is, this step can also be detected and executed after the Nth round of task allocation of the parallel detection channel. Specifically, the visual recognition system first verifies the reliability of the detected anomaly, including multiple repeated detections, cross-validation, threshold comparison, etc.; then evaluates the severity of the anomaly, including calculating the anomaly parameters, analyzing the scope of influence, and determining the anomaly level; then sends a stop command to all parallel detection channels to ensure that each channel stops the current detection task in an orderly manner; at the same time, saves the intermediate status and completed detection results of all detection channels; finally, generates an anomaly detection report, recording complete information such as the time, location, characteristics, and severity of the anomaly discovery.

[0079] In some embodiments, anomaly detection and processing can be achieved through a variety of methods: Optionally, the visual recognition system can establish a multi-level anomaly detection model and identify complex anomaly patterns through deep learning algorithms, specifically including: feature extraction, pattern matching, anomaly scoring, threshold judgment, result verification, and anomaly confirmation; Optionally, the visual recognition system can use statistical process control methods to establish anomaly discrimination criteria based on historical data, specifically including: data collection, statistical analysis, control limit calculation, trend monitoring, anomaly identification, and result output. It is understood that other intelligent detection or anomaly identification methods can also be used to achieve accurate judgment of process anomalies, which are not limited here.

[0080] During exception handling, false anomalies can lead to unnecessary detection interruptions. To address this, the visual recognition system employs multiple error prevention mechanisms: It establishes an anomaly signature library to distinguish common anomalies from interference factors; establishes a multi-level anomaly confirmation process, requiring high-confidence anomalies to trigger a halt; implements a rapid recovery mechanism to allow for rapid resumption of detection in the event of a false alarm; and maintains a complete basis for decision-making to support later analysis of anomaly judgments. For example, when a detection channel reports an anomaly, the system immediately initiates cross-validation, confirming the anomaly's authenticity through other detection methods or historical data comparisons. A halt is only executed if multiple verifications confirm the presence of an anomaly.

[0081] In the above example, process correlation analysis was used to determine inspection priorities, and multi-channel parallel inspection was employed to improve processing efficiency. In actual applications, the system will also dynamically update the inspection sequence based on real-time inspection results to ensure optimal allocation of inspection resources. The following supplements the scenarios of this example.

[0082] In further applications, the visual recognition system can also combine real-time parameters of production equipment to provide early warnings. For example, when parameters such as temperature and pressure of surface treatment equipment fluctuate, the system automatically records the relevant product numbers and focuses on the surface quality of these products in subsequent inspections. Simultaneously, the system continuously accumulates empirical data based on test results to optimize the calculation model for process correlation. For processes that pass inspection, the system records their process parameter combinations to establish the optimal process window. For processes with abnormalities, the system analyzes the causes and automatically generates parameter adjustment recommendations. This intelligent quality control system significantly improves production efficiency and product yield.

[0083] After combining the above scenarios, the following is a more detailed description of the process of the method provided by this implementation. Figure 2 , is another flow chart of the method for visually identifying good products in an embodiment of the present application.

[0084] S201. Obtain product monitoring data from multiple monitoring devices in a product production process.

[0085] The product production process represents the complete manufacturing process of a laptop computer case, from raw materials to finished product, including multiple steps such as injection molding, machining, surface treatment, and painting. Monitoring equipment refers to data acquisition devices installed at each step of the production line, including image collectors, temperature sensors, and pressure sensors. Product monitoring data represents information such as process parameters, image data, and quality indicators collected in real time by each monitoring device, reflecting the real-time status of the product production process. Multiple monitoring devices represent a collection of various types of data acquisition devices deployed at different steps and locations.

[0086] Visual recognition systems need to continuously acquire monitoring data from the production process to achieve real-time monitoring of product quality. Specifically, the visual recognition system first determines the acquisition parameters and data format of each monitoring device; then establishes a data communication connection with each device, sets the data acquisition frequency and caching strategy; then, according to pre-set acquisition rules, it acquires monitoring data from each device, including image data and process parameters; simultaneously, it pre-processes the collected data, including data cleaning, format conversion, and time synchronization; finally, it stores the processed data in a database in a unified format and creates an index to facilitate subsequent analysis.

[0087] In some embodiments, the collection and processing of monitoring data can be achieved through a variety of methods: Optionally, the visual recognition system can adopt a distributed data collection architecture to acquire monitoring data by deploying local collection nodes in each process and centrally aggregating the data, specifically including: configuring the collection parameters of the local collection nodes, establishing a communication network between the nodes, implementing data collection, local preprocessing, data transmission, and central aggregation; Optionally, the visual recognition system can build a data collection system based on the industrial Internet of Things platform, specifically including: connecting to Internet of Things devices, configuring collection protocols, establishing data models, collecting data, real-time processing, and cloud storage. It is understandable that other data collection and processing methods can also be used to achieve the acquisition of monitoring data, which is not limited here.

[0088] During the data collection process, data loss may occur due to device communication interruptions. To address this, the visual recognition system employs multiple fault-tolerance mechanisms: a backup channel for data collection is established, automatically switching to the backup channel when the primary channel fails; a data cache is set up on the local node to temporarily store data during network outages and automatically retransmit it upon recovery; data integrity verification rules are set to interpolate and repair lost data; and a data anomaly log is recorded to support retrospective analysis of the data collection process. For example, if communication with the image acquisition device in a certain process is interrupted, the system automatically switches to the backup camera to continue collection, while temporarily storing the process parameter data during the interruption locally for retransmission and verification after communication is restored.

[0089] S202: Perform feature recognition on the product monitoring data, identify marked products whose product feature values ​​deviate from a preset threshold range, and generate abnormal prompt information.

[0090] Feature recognition involves analyzing and processing monitoring data to extract key features that reflect product quality. Product feature values ​​refer to various parameters that characterize product quality, including quantitative indicators such as size, appearance, and performance. Preset thresholds represent the normal fluctuation range of each feature value and are used to determine whether a product is abnormal. Flagged products are product instances identified as having abnormal feature values. Abnormality information records and displays specific information about abnormal products, including the type of abnormality, severity of the abnormality, and time of occurrence.

[0091] The visual recognition system needs to analyze collected monitoring data in real time to promptly identify abnormal products. Specifically, the visual recognition system first reads the latest monitoring data from the database, including image data and process parameters. It then preprocesses the data, including denoising, enhancement, and standardization. It then uses a feature extraction algorithm to analyze the data and calculate various eigenvalues. These calculated eigenvalues ​​are compared with preset thresholds. If a eigenvalue exceeds the threshold, the system records the relevant product information and determines an alarm level based on the severity of the anomaly. Finally, it generates a notification message containing details of the anomaly and sends it to the relevant personnel.

[0092] In some embodiments, feature recognition and anomaly detection can be implemented in a variety of ways: Optionally, the visual recognition system can employ a deep learning model for feature extraction and anomaly detection, specifically including image preprocessing, feature extraction, pattern recognition, anomaly scoring, threshold determination, and result output. Optionally, the visual recognition system can construct an anomaly detection framework based on statistical process control methods, specifically including data acquisition, statistical analysis, control limit calculation, trend monitoring, rule determination, and anomaly labeling. It is understood that other feature analysis and anomaly detection methods may also be employed, and are not limited here.

[0093] During the feature recognition process, false positives and missed negatives can be problematic. To address this, the visual recognition system employs a multi-verification mechanism: establishing a multi-level threshold system with different handling strategies for varying degrees of anomaly; introducing temporal correlation analysis to combine historical data to determine the reliability of anomalies; setting cross-validation rules to improve accuracy through the combination of multiple features; and recording the basis for judgment to support later review of anomaly judgments. For example, when a suspected scratch is detected on a product's surface, the system simultaneously analyzes multiple dimensions, including regional grayscale values, edge features, and texture features, and combines the judgment results of similar cases in historical data to ultimately determine whether to mark the product as an anomaly.

[0094] S203: After a preset waiting time, or when a product tracing instruction is received from the user after viewing the abnormal prompt information, a product tracing application is generated.

[0095] The preset waiting time indicates the interval at which the system automatically triggers a traceability request, automatically initiating the traceability process without human intervention. Exception information refers to system-generated warnings about abnormal products. Product traceability instructions indicate traceability testing commands issued after the user confirms and reviews the exception information. A product traceability request is a formal request to initiate the traceability testing process, including the exception information, trigger method, and timestamp.

[0096] The visual recognition system needs to choose the appropriate time to generate a traceability request based on different scenarios. Specifically, the system first sets a preset waiting time to control the interval between automatic triggering. When the system generates an exception prompt, it starts the waiting timer. During the waiting period, if the user's traceability instruction is received, the traceability request is immediately generated. If the user instruction is not received after the preset waiting time, the traceability request is automatically generated. The triggering method, time point, and related exception information of the traceability request are also recorded. The traceability request is stored in the database and the relevant modules are notified to initiate the detection process.

[0097] In some embodiments, the generation and management of traceability applications can be achieved through a variety of methods: Optionally, the visual recognition system can use a state machine mechanism to manage the generation process of traceability applications, specifically including: anomaly detection, wait timer, trigger judgment, application generation, state transition, and process control; Optionally, the visual recognition system can build an application triggering framework based on an event-driven model, specifically including: event monitoring, condition judgment, priority assessment, resource inspection, application packaging, and task dispatching. It is understood that other process management or task scheduling methods can also be used, which are not limited here.

[0098] During the traceability application generation process, a large number of exceptions within a short period of time can cause a backlog of applications. To address this, the visual recognition system employs an intelligent scheduling mechanism: It establishes a traceability application priority assessment model to determine the processing order based on factors such as the degree of anomaly and the scope of impact; sets application merging rules to batch process anomalies within the same process and at similar time points; implements a resource pre-check mechanism to dynamically adjust the automatic triggering interval based on system load; and records application processing logs to support optimized analysis of the traceability process. For example, when similar surface defects appear on multiple products simultaneously, the system performs cluster analysis on these anomalies and generates a batch traceability application, improving processing efficiency.

[0099] S204: In response to the product traceability application, determine the test type of the product to be traced.

[0100] Referring to step S101 , the visual recognition system determines the detection type.

[0101] It should be noted that determining the inspection type involves a complex decision-making and analysis process. The visual recognition system uses a hierarchical decision-making model to construct an inspection needs assessment function by comprehensively evaluating factors such as the anomaly type characteristics, the product's production stage, and historical inspection records. Based on the evaluation results, the system also calculates the suitability scores for both process inspection and final inspection, automatically selecting the type with the higher score. For example, when a localized scratch is detected on the surface of the casing, the system analyzes the defect's formation characteristics, location distribution, and historical cases. If it finds that similar defects are commonly caused by a specific process, it will tend to select the process inspection type.

[0102] S205: When the detection type is a process detection request, obtain abnormal process information that triggers the detection.

[0103] Referring to step S102 , the visual recognition system determines abnormal process information when a process inspection request is made.

[0104] S206 , calculating the process correlation between each completed process and the abnormal process, and sorting each completed process based on the process correlation to generate a sequence to be detected.

[0105] Referring to step S103 , the visual recognition system generates a sequence to be detected.

[0106] It should be noted that the process correlation calculation can use an improved Markov random field model to represent the correlation relationship between processes as a conditional probability network. The visual recognition system establishes the conditional influence probability and process weight matrix between processes, analyzes the process parameter transmission chain and the evolution law of quality characteristics, and constructs a complete correlation evaluation system. In actual applications, the system will consider the direct influence and indirect transmission effects between processes, and dynamically adjust the correlation calculation parameters based on the abnormal correlation patterns in historical data. For example, during the shell processing process, if an abnormality occurs in the finishing process, the system will analyze the correlation of the previous rough processing process and consider the influence weights of factors such as tooling and processing parameters.

[0107] S207: When the inspection type is a final inspection request, extract defect feature data of the product to be traced.

[0108] Referring to step S104 , the visual recognition system extracts defect feature data when a final inspection is requested.

[0109] S208. Based on the defect feature data, determine a set of related processes from the process flow database, and sort them to generate a sequence to be inspected.

[0110] Referring to step S105 , the visual recognition system generates a sequence to be detected.

[0111] It should be noted that defect feature data analysis uses a multimodal feature fusion approach. The feature expression model constructed by the visual recognition system includes multiple dimensions, including visual features, texture features, and position features. These features are extracted through a deep learning network and a mapping relationship is established with process parameters, thereby tracing possible problematic processes. This mapping relationship takes into account the importance weights of different feature dimensions and continuously optimizes feature extraction and mapping rules based on successful cases in historical data. For example, when the final inspection reveals a bubble defect on the shell surface, the system will analyze the bubble's morphological characteristics, spatial distribution characteristics, and depth characteristics, and locate the possible problematic process through the feature-process mapping relationship.

[0112] S209 , starting multiple parallel inspection channels, and sequentially acquiring images of the processes to be inspected that rank top in the sequence to be inspected, and inputting the images of the processes to be inspected into the parallel inspection channels for inspection.

[0113] Referring to step S106 , the visual recognition system performs recognition detection.

[0114] It should be noted that the scheduling of parallel inspection channels utilizes a dynamic load balancing algorithm. The visual recognition system establishes a resource allocation model by evaluating the complexity of inspection tasks and resource usage. This model optimizes the allocation of inspection resources by minimizing load differences between channels. The system also implements an adaptive task allocation strategy that comprehensively considers factors such as current load conditions, inspection quality requirements, and processing priorities. For example, when multiple processes need to be inspected simultaneously, the system intelligently allocates inspection tasks based on the image feature complexity and processor load of each process.

[0115] In some embodiments, the visual recognition system will allocate detection tasks according to the hardware performance of each parallel detection channel and make rational use of detection resources. That is, the visual recognition system will start multiple parallel detection channels and obtain the hardware parameter information corresponding to each parallel detection channel; perform computing power grading on each parallel detection channel according to the hardware parameter information to obtain computing power grading results; obtain the images of the processes to be inspected of the top-ranked processes in the sequence to be inspected in turn, and calculate the estimated detection time of the images of the processes to be inspected; based on the computing power grading results and the estimated detection time, perform channel allocation on the images of the processes to be inspected.

[0116] Hardware parameter information represents the computing performance indicators of parallel inspection channels, including processor speed, memory capacity, cache size, and more. Computing power grading represents the classification of inspection channel processing capabilities. Estimated inspection time is the estimated time required to complete process image inspection. Channel allocation refers to the rational allocation of inspection resources based on task requirements and channel capabilities.

[0117] The visual recognition system must perform resource assessment and task allocation before initiating inspection. Specifically, the system first obtains hardware configuration information for all inspection channels; then, it assesses the channels based on performance metrics; analyzes the feature complexity of each process image in the inspection sequence; estimates the inspection time based on image features; analyzes the matching of inspection tasks with channel capabilities; develops an optimal task allocation plan; and reserves some high-performance channels for urgent tasks. Finally, it executes task allocation and initiates the inspection process. The system continuously monitors inspection progress and makes dynamic adjustments when necessary.

[0118] In some embodiments, resource assessment and task allocation can be achieved through a variety of methods: Optionally, the visual recognition system can use performance modeling methods for resource management, including hardware performance assessment, load capacity calculation, resource pool division, task demand analysis, matching calculation, and allocation strategy generation; Optionally, the visual recognition system can optimize the detection process based on a task scheduling algorithm, including task decomposition, dependency analysis, priority assessment, resource reservation, load balancing, and real-time scheduling. It is understood that other resource management or scheduling algorithms can also be used to achieve improved detection efficiency, and this is not limited here.

[0119] During resource allocation, uneven task load distribution can lead to low resource utilization. To address this, the visual recognition system employs a dynamic load balancing mechanism: a real-time load monitoring system is established to track resource usage across channels; a dynamic task migration strategy is implemented to flexibly allocate tasks between channels; load thresholds are set to prevent overloading of individual channels; and resource utilization data is recorded to support optimization and adjustment of allocation strategies. For example, if an overloaded channel is detected during inspection, the visual recognition system will dynamically migrate some tasks to a less-loaded channel, ensuring optimal overall inspection efficiency.

[0120] S210 : When the parallel inspection channel completes the inspection of the current process and the result is passed, generate inspection pass data based on the current process.

[0121] Parallel inspection channels represent multiple processing units performing image inspection simultaneously. The current process refers to the production step currently undergoing inspection. A pass result indicates that the product quality of the current process meets standard requirements. Pass data is a complete set of information recording the pass status of an inspection, including process information, inspection parameters, and confidence level.

[0122] The visual recognition system needs to promptly process the results of inspections and generate corresponding data. Specifically, the system first verifies the reliability of the inspection results, including checking whether the inspection parameters are within the valid range and whether the inspection process is stable and complete. It then records detailed information about the inspection results, including the process ID, inspection time, inspection parameters, and image features. It then calculates the confidence level of the inspection results, taking into account factors such as inspection conditions and parameter stability. The inspection data is packaged in a standard format. The process status is updated, marking the process as completed. Finally, the inspection data is stored in a database for subsequent analysis and optimization.

[0123] In some embodiments, the generation and management of detection data can be achieved through a variety of methods: Optionally, the visual recognition system can use a multidimensional feature analysis method to evaluate detection results, specifically including: parameter verification, feature extraction, stability analysis, confidence calculation, data encapsulation, and result storage; Optionally, the visual recognition system can construct a detection result evaluation system based on a knowledge graph, specifically including: feature mapping, rule matching, association analysis, reliability assessment, data generation, and status update. It is understood that other data analysis or result evaluation methods can also be used, and are not limited here.

[0124] During the generation of pass data, unstable test results can make confidence assessment difficult. To address this, the visual recognition system employs a dynamic assessment mechanism: a multi-dimensional stability assessment model is established, taking into account test parameters, environmental factors, equipment status, and more. Dynamic threshold rules are set to adaptively adjust judgment criteria based on historical data. Time series analysis is introduced to assess trend changes in test results. The assessment process is recorded to support retrospective analysis of the reliability of the results. For example, when the test results of a process fluctuate significantly, the system increases the number of samples and analyzes the causes of the fluctuations based on historical data to ensure sufficient reliability of the generated pass data.

[0125] In some embodiments, the visual recognition system will summarize the inspection results and record the information of passed inspection in the inspection pass data to make the optimization more accurate. That is, the visual recognition system will update the inspection pass data based on the corresponding current process inspection when the corresponding current process inspection is completed in other parallel inspection channels and the result is passed.

[0126] Parallel inspection channels represent multiple independent processing units that simultaneously perform inspection tasks, improving inspection efficiency. The current process refers to the production link undergoing inspection. A pass result indicates that the process quality inspection meets the preset standard requirements. Pass inspection data records the inspection information of qualified processes, including process parameters, inspection time, and test results. An update operation adds new inspection results to the existing pass inspection data set.

[0127] The visual recognition system needs to promptly update the inspection record when any parallel inspection channel completes inspection and the result is qualified. Specifically, the visual recognition system first confirms the inspection status of the inspection channel and verifies the integrity of the inspection process; then conducts a reliability assessment of the inspection results to ensure the accuracy of the inspection data; then extracts the inspection parameters and result data of the current process; encapsulates the new inspection data in a standard format; updates the relevant records in the inspection pass database; and simultaneously updates the process status mark to indicate that the process inspection is completed; finally, triggers the subsequent data analysis and optimization process.

[0128] In some embodiments, the update processing of detection pass data can be implemented in a variety of ways: Optionally, the visual recognition system can use a distributed data synchronization mechanism to update detection records, including data verification, incremental updates, state synchronization, consistency checks, version management, and logging; Optionally, the visual recognition system can manage the data update process based on an event-driven model, including event triggering, data encapsulation, update strategy formulation, data merging, integrity verification, and state push. It is understood that other data management or update methods can also be used to implement the update of detection pass data, which is not limited here.

[0129] S211 : Based on the passed detection data, the associated processes in the sequence to be detected are removed to obtain the optimized sequence to be detected.

[0130] Among them, "passed inspection data" represents complete inspection information for processes that have been confirmed to meet quality standards. "Dependent processes" refer to processes that have a process dependency or influence on a process that has passed inspection. "To-be-inspected sequence" represents a queue of processes that have not yet completed inspection. "Optimized to-be-inspected sequence" refers to a queue of processes that have undergone optimization and adjustment.

[0131] The visual recognition system needs to dynamically optimize the sequence to be inspected based on the inspection results. Specifically, the system first analyzes the process characteristics of the processes that have passed the inspection and identifies their key impact on the production process. It then extracts the correlation between processes from the process database and constructs an influence transmission network. It calculates the degree of correlation between other processes to be inspected and the current process. It analyzes processes with correlations exceeding a threshold to assess whether they can be removed from the sequence to be inspected. It then generates a new optimized sequence to ensure the completeness of the inspection coverage. The optimization results are recorded in the database for subsequent inspection optimization.

[0132] In some embodiments, sequence optimization can be achieved through a variety of methods: Optionally, the visual recognition system can employ a process dependency graph analysis method, specifically including: constructing a dependency network, calculating influence strength, determining optimization strategies, sequence adjustment, integrity verification, and result updating; Optionally, the visual recognition system can construct a sequence optimization model based on a heuristic algorithm, specifically including: feature extraction, correlation analysis, optimization target setting, sequence reconstruction, verification and evaluation, and dynamic updating. It is understood that other optimization algorithms or decision-making methods may also be employed, and are not limited herein.

[0133] During sequence optimization, over-optimization can lead to missed detections. To address this, the visual recognition system employs a conservative optimization strategy: establishing a process impact assessment model to strictly control the conditions for process removal; setting up multi-level verification rules to ensure that optimization does not affect detection reliability; retaining key node processes, even if they are correlated, in the detection sequence; and recording the basis for optimization to support retrospective analysis of sequence adjustments. For example, when the system determines that certain processes can be removed, it first evaluates the historical anomaly probability of these processes and only removes them when the anomaly probability is extremely low and they are highly correlated with processes that have already passed inspection.

[0134] In some embodiments, the visual recognition system will perform a reliability assessment on the detection results and initiate a re-inspection mechanism for results with low confidence levels. That is, the visual recognition system will first obtain the confidence value of the detection pass data of the current process; then, when the confidence value is lower than a preset confidence threshold, the first process in the sequence to be detected will be obtained and sent to the parallel detection channel for detection, and the current process will be added to the first position in the sequence to be detected.

[0135] The confidence value represents a quantitative indicator of the reliability of the test result and is used to assess the confidence level of the test quality. The preset confidence threshold is the standard value for determining the reliability of the test result. The first position in the queue for testing represents the highest priority process to be tested. Adding to the first position moves the process to the highest priority position in the queue for testing.

[0136] The visual recognition system needs to evaluate the reliability of the inspection results and trigger re-inspection if the reliability is insufficient. Specifically, the visual recognition system first extracts the confidence index of the current process from the inspection data; then compares the confidence value with a preset threshold; when the confidence value is lower than the threshold, the current process is marked as requiring re-inspection; then the process is added to the front of the sequence to be inspected; at the same time, the original first process is retrieved from the sequence to be inspected; the inspection tasks of the two processes are respectively assigned to the appropriate inspection channels; a new inspection process is initiated; and finally, the process status and inspection plan are updated.

[0137] In some embodiments, confidence assessment and re-inspection processing can be achieved in a variety of ways: Optionally, the visual recognition system can adopt a multi-dimensional confidence assessment method, including feature matching analysis, detection stability assessment, environmental impact assessment, historical data comparison, comprehensive score calculation, and decision threshold judgment; Optionally, the visual recognition system can manage the detection process based on an adaptive re-inspection strategy, including abnormality assessment, resource status check, priority recalculation, channel reservation, task reallocation, and status synchronization. It is understandable that other assessment methods or re-inspection strategies can also be used to achieve detection quality assurance, which is not limited here.

[0138] During the confidence assessment and re-inspection process, repeated inspections can lead to decreased inspection efficiency. To address this, the visual recognition system employs an intelligent re-inspection optimization mechanism: It establishes an adaptive inspection parameter adjustment strategy to optimize re-inspection parameters based on the initial inspection results; implements an inspection resource reservation mechanism to ensure the timely execution of re-inspection tasks; sets a maximum re-inspection limit to avoid ineffective repetitions; and records re-inspection reason analysis to support continuous optimization of inspection strategies. For example, when the confidence level of the initial inspection of a process is insufficient, the visual recognition system analyzes the specific factors affecting the confidence level and adjusts the inspection parameters or switches the inspection method accordingly to improve the re-inspection success rate.

[0139] S212: re-sort the optimized sequence to be inspected, and obtain the sorted first process image to the parallel inspection channel for inspection.

[0140] Reordering refers to adjusting the inspection order of processes within the optimized sequence based on a specific strategy. The top process is the process with the highest priority after sorting. Process images represent product image data corresponding to the process to be inspected. Parallel inspection channels refer to a collection of multiple processing units that can simultaneously execute image inspection tasks.

[0141] The visual recognition system needs to scientifically and rationally arrange the inspection sequence and allocate inspection resources. Specifically, the visual recognition system first analyzes the characteristics of each process in the optimized sequence to be inspected, including process type, abnormality risk, and inspection difficulty. It then calculates the inspection priority of each process based on multi-dimensional evaluation indicators, taking into account factors such as process importance, resource consumption, and time constraints. The sequence is sorted based on priority to generate the final inspection order. Inspection resources are allocated to the first process after sorting, including determining the inspection channel to be used and loading inspection parameters. The image data of the first process is retrieved from the image database. The image data is distributed to the designated parallel inspection channels, initiating the inspection process.

[0142] In some embodiments, detection sorting and resource allocation can be achieved through a variety of methods: Optionally, the visual recognition system can use a multi-objective optimization algorithm to make sorting decisions, specifically including: feature quantification, weight calculation, priority assessment, sequence rearrangement, resource allocation, and task dispatching; Optionally, the visual recognition system can construct a detection scheduling model based on a dynamic programming method, specifically including: state assessment, objective function construction, optimal path solution, sequence generation, resource scheduling, and task initiation. It is understood that other scheduling algorithms or resource allocation methods can also be used, and are not limited here.

[0143] During the sorting and inspection allocation process, an imbalance in the inspection resource load can occur. To address this, the visual recognition system employs an adaptive scheduling mechanism: It establishes an inspection channel load assessment model to monitor resource usage in real time across each channel; sets load balancing rules to dynamically adjust task allocation strategies based on channel status; introduces a task estimation mechanism to plan resource allocation plans in advance; and records the scheduling process to support optimized analysis of resource utilization efficiency. For example, when the processing queue for a particular inspection channel is long, the system will select a channel with a lighter load and suitable for that type of inspection based on the inspection characteristics of the first process, thereby achieving rational resource utilization. Furthermore, the system dynamically adjusts the parameters of the priority calculation model based on feedback from inspection results to continuously optimize the sorting strategy.

[0144] S213. When a parallel detection channel detects a process abnormality, the corresponding process is determined as an abnormal process, and detection of all parallel detection channels is stopped.

[0145] Referring to step S108, the visual recognition system will stop inspection if it determines that a process anomaly exists. Similarly, this step can be implemented at any time. That is, step S213 can be performed after step S209, i.e., after the first round of task allocation for the parallel inspection channel, or after step S212, i.e., after the second round of task allocation for the parallel inspection channel. It can be easily deduced that steps S210 to S212 can be multiple, i.e., this step can also be performed after the Nth round of task allocation for the parallel inspection channel.

[0146] It should be noted that the termination judgment of anomaly detection can adopt a rapid response mechanism. The anomaly propagation model constructed by the visual recognition system takes into account multiple dimensions such as the evolution of defect characteristics, detection confidence, and impact range, realizing real-time assessment of abnormal conditions. This model can quickly analyze the reliability and potential impact of anomalies and trigger termination instructions when it is confirmed that the impact of the anomaly exceeds the allowable range. For example, when a critical dimension of a process is detected to be out of tolerance, the system will evaluate the impact of the anomaly on subsequent processes, analyze the risk of anomaly propagation, and promptly terminate the operation of all detection channels after confirming the reliability of the anomaly.

[0147] In some embodiments, the visual recognition system will perform intelligent analysis of abnormal situations and provide processing suggestions to optimize and adjust production equipment. That is, the visual recognition system will obtain detection images and detection parameters of abnormal processes and generate an abnormal analysis report; generate abnormal detection results and parameter adjustment suggestions based on the abnormal analysis report, and receive process adjustment instructions input by the user; and adjust the process parameters of production equipment for abnormal processes according to the process adjustment instructions.

[0148] The "Abnormality Analysis Report" provides a detailed record and analysis of abnormalities detected during testing. "Parameter Adjustment Suggestions" provide process parameter optimization solutions based on abnormality analysis. "Process Adjustment Instructions" provide specific parameter adjustment requirements after user confirmation. "Production Equipment" refers to the specific equipment used to perform the process.

[0149] After detecting an anomaly, the visual recognition system needs to analyze and diagnose it and guide process adjustments. Specifically, the system first collects inspection images and parameter data of the abnormal process; then extracts and analyzes the abnormal features; generates an analysis report containing an abnormality description, cause analysis, and impact assessment; formulates parameter adjustment recommendations based on the cause of the abnormality; provides the analysis results and adjustment recommendations to the user; receives and verifies the user's adjustment instructions; and finally executes the process parameter adjustments, while simultaneously monitoring the production status after the adjustments.

[0150] In some embodiments, anomaly analysis and parameter adjustment can be achieved through a variety of methods: Optionally, the visual recognition system can use knowledge graph methods to perform anomaly diagnosis, including feature extraction, pattern matching, causal analysis, impact assessment, solution generation, and verification and evaluation; Optionally, the visual recognition system can optimize parameter adjustment based on a feedback control model, including state assessment, target setting, deviation analysis, parameter calculation, adjustment execution, and effect verification. It is understood that other analysis methods or optimization strategies can also be used to achieve process improvement, and are not limited here.

[0151] During the parameter adjustment process, unstable adjustment results can be a problem. To address this, the visual recognition system employs a progressive adjustment strategy: establishing a parameter sensitivity analysis model to assess the impact of each parameter adjustment; implementing a step-by-step adjustment mechanism to achieve optimization goals through multiple small adjustments; setting parameter adjustment range limits to ensure adjustment safety; and recording adjustment process data to support continuous optimization of the adjustment strategy. For example, when adjusting the temperature parameters of an injection molding process, the visual recognition system will first conduct a small-scale trial adjustment and then, based on feedback, gradually optimize until the desired improvement is achieved.

[0152] In the embodiment of the present application, due to the adoption of a dual-path detection strategy based on process correlation and a multi-channel parallel detection mechanism, it is possible to intelligently analyze possible problem processes based on abnormal characteristics, and simultaneously carry out detection through parallel channels, effectively solving the problems of single detection paths and low serial processing efficiency in traditional technologies, thereby achieving efficient utilization of detection resources and rapid location of problems. At the same time, the system also has functions such as real-time monitoring and early warning, dynamic sequence optimization, and abnormal analysis and diagnosis, which can promptly detect potential problems and provide improvement suggestions, significantly improving the automation level of the production line and product quality control capabilities. Through this intelligent visual recognition method, not only is the efficiency of problem location improved and the waste of detection resources reduced, but it also provides data support for process optimization and achieves continuous improvement of the production process.

[0153] The visual recognition system in the embodiment of the present invention is described below from the perspective of hardware processing. Figure 3 , is a schematic diagram of the physical device structure of the visual recognition system in an embodiment of the present application.

[0154] It should be noted that Figure 3 The structure of the visual recognition system shown is only an example and should not limit the functions and scope of use of the embodiments of the present invention.

[0155] like Figure 3 As shown, the visual recognition system includes a CPU 301, which can perform various appropriate actions and processes according to the programs stored in the ROM 302 or the programs loaded from the storage unit 308 into the RAM 303, such as executing the methods described in the above embodiments. The RAM 303 also stores various programs and data required for system operation. The CPU 301, ROM 302, and RAM 303 are connected to each other via a bus 304. An I / O interface 305 is also connected to the bus 304.

[0156] The following components are connected to the I / O interface 305: an input section 306 including an audio input device, push button switches, and the like; an output section 307 including a liquid crystal display (LCD), an audio output device, indicator lights, and the like; a storage section 308 including a hard disk and the like; and a communication section 309 including a network interface card such as a LAN (Local Area Network) card or a modem. The communication section 309 performs communication processing via a network such as the Internet. A drive 310 is also connected to the I / O interface 305 as needed. Removable media 311, such as a magnetic disk, an optical disk, a magneto-optical disk, or a semiconductor memory, is installed in the drive 310 as needed, so that computer programs read from the removable media can be installed in the storage section 308 as needed.

[0157] In particular, according to an embodiment of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, an embodiment of the present invention includes a computer program product comprising a computer program carried on a computer-readable medium, the computer program including a computer program for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network via the communication section 309 and / or installed from the removable medium 311. When the computer program is executed by the CPU 301, the various functions defined in the present invention are performed.

[0158] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present invention. Each box in the flowchart or block diagram can represent a module, program segment, or part of the code, and the above-mentioned module, program segment, or part of the code contains one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in an order different from that marked in the accompanying drawings.

[0159] Specifically, the visual recognition system of this embodiment includes a processor and a memory. The memory stores a computer program. When the computer program is executed by the processor, the visual recognition method for good products provided in the above embodiment is implemented.

[0160] As another aspect, the present invention further provides a computer-readable storage medium, which may be included in the visual recognition system described in the above embodiments, or may exist independently and not be incorporated into the visual recognition system. The storage medium carries one or more computer programs, which, when executed by a processor of the visual recognition system, enable the visual recognition system to implement the method for visually identifying defective products provided in the above embodiments.

[0161] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

[0162] As used in the above embodiments, the term “when” may be interpreted to mean “if” or “after” or “in response to determining that” or “in response to detecting that”, depending on the context. Similarly, the phrases “upon determining that” or “if (stated condition or event) is detected” may be interpreted to mean “if determining that” or “in response to determining that” or “upon detecting (stated condition or event)” or “in response to detecting (stated condition or event)”, depending on the context.

Claims

1. A method for visually identifying good products, characterized in that: Applied to a visual recognition system, the method includes: In response to a product traceability application, determining a test type for the product to be traced; the test type includes a process test request or a final inspection request; When the detection type is the process detection request, obtaining abnormal process information that triggers the detection; Calculating the process correlation between each completed process and the abnormal process, and sorting the completed processes based on the process correlation to generate a sequence to be detected; the process correlation represents the direct correlation between processes, the correlation between process parameters, and the transfer relationship of quality characteristics; the process correlation calculation steps include: extracting the topological relationship between processes, the process parameter influence matrix, and the quality characteristic transfer chain from the process flow database; calculating the influence weight of the completed process on the abnormal process, the influence weight including direct influence, indirect influence, and cumulative influence, and introducing a time series attenuation factor; using a graph theory algorithm or a multi-factor weighted scoring method to calculate the process correlation; When the inspection type is the final inspection request, extracting defect feature data of the product to be traced; Based on the defect feature data, a set of related processes is determined from a process flow database, and the process steps are sorted to generate a sequence to be inspected; Starting multiple parallel detection channels, and sequentially acquiring images of the processes to be inspected that rank top in the sequence to be inspected, and inputting the images of the processes to be inspected into the parallel detection channels for inspection; When the parallel inspection channel completes the inspection of the current process and the result is passed, the sequence to be inspected is updated, and the image of the first process in the updated sequence to be inspected is obtained for inspection. The updating step of the sequence to be inspected includes: analyzing the association relationship between the current process and other processes; updating the association weight and recalculating the inspection priority, wherein the inspection priority is adjusted using the process dependency network model and the graph structure update algorithm; and generating an updated sequence to be inspected. When the parallel detection channels detect that a process is abnormal, the corresponding process is determined as an abnormal process, and detection of all the parallel detection channels is stopped.

2. The method according to claim 1, characterized in that Before the step of determining the inspection type of the product to be traced in response to the product traceability application, the method further includes: Obtain product monitoring data from multiple monitoring devices in the product production process; Performing feature recognition on the product monitoring data, identifying marked products whose product feature values ​​deviate from a preset threshold range, and generating abnormal prompt information; After a preset waiting time, or when a product tracing instruction input by the user after viewing the abnormal prompt information is received, a product tracing application is generated.

3. The method according to claim 1, characterized in that When the parallel inspection channel completes the inspection of the current process and the result is passed, the step of updating the sequence to be inspected and obtaining the first process image in the updated sequence to be inspected for inspection specifically includes: When the parallel inspection channel completes the inspection of the current process and the result is passed, generating inspection pass data based on the current process; Based on the passed detection data, removing associated processes in the sequence to be detected to obtain an optimized sequence to be detected; The optimized sequence to be inspected is reordered, and the image of the first process after the order is acquired and sent to the parallel inspection channel for inspection.

4. The method according to claim 3, characterized in that When the parallel inspection channel completes the inspection of the current process and the result is passed, after the step of generating inspection pass data based on the current process, the method further includes: When other parallel inspection channels complete the corresponding current process inspection and the result is passed, the inspection pass data is updated based on the corresponding current process inspection.

5. The method according to claim 3, characterized in that Before the step of removing associated processes in the sequence to be inspected based on the passed inspection data to obtain the optimized sequence to be inspected, the method further includes: Obtaining a confidence value of the inspection pass data of the current process; When the confidence value is lower than a preset confidence threshold, the first process of the sequence to be detected is obtained and sent to the parallel detection channel for detection, and the current process is added to the first position of the sequence to be detected.

6. The method according to claim 1, characterized in that The step of starting multiple parallel detection channels, sequentially acquiring images of the processes to be inspected that rank top from the sequence to be inspected, and inputting the images of the processes to be inspected into the parallel detection channels for inspection specifically includes: Starting multiple parallel detection channels and obtaining hardware parameter information corresponding to each of the parallel detection channels; Performing computing power classification on each of the parallel detection channels according to the hardware parameter information to obtain a computing power classification result; Sequentially acquiring images of the processes to be inspected that rank top in the sequence to be inspected, and calculating an estimated inspection time for the images of the processes to be inspected; Based on the computing power classification result and the estimated detection time, channels are allocated to the process images to be inspected.

7. The method according to claim 1, characterized in that After the step of determining the corresponding process as an abnormal process and stopping detection of all the parallel detection channels when the parallel detection channels detect that the process is abnormal, the method further includes: Acquire detection images and detection parameters of the abnormal process and generate an abnormality analysis report; Generate abnormality detection results and parameter adjustment suggestions based on the abnormality analysis report, and receive process adjustment instructions input by the user; Adjust the process parameters of the production equipment of the abnormal process according to the process adjustment instruction.

8. A visual recognition system, characterized in that: The visual recognition system includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to enable the visual recognition system to execute the method described in any one of claims 1-7.

9. A computer-readable storage medium comprising instructions, characterized in that: When the instruction is executed on a visual recognition system, the visual recognition system is caused to execute the method according to any one of claims 1 to 7.

10. A computer program product, characterized in that When the computer program product is run on a visual recognition system, the visual recognition system is caused to perform the method according to any one of claims 1 to 7.

Images