Detection and identification method and system based on artificial intelligence

Abstract

The invention discloses a detection and identification method and system based on artificial intelligence, and relates to the related technical field of image data processing, and the method comprises the steps: calling a fault record and mining a fault feature tree for a target automobile part; and based on the fault feature tree, taking multi-stage hash analysis based on a pixel array as a detection mode, performing sample driving training, and constructing a defect detection model. By controlling front-end sensing equipment, image acquisition is carried out on a target automobile part, and a part image is determined. And returning the part image to the defect detection model, performing part defect identification of the pixel array under iterative sampling, and outputting a detection chain. And based on the detection chain, processing guidance of the target automobile part is carried out. The technical problems that in the prior art, an existing automobile part defect detection method is low in detection precision and efficiency, and complex and changeable defect types are difficult to deal with are solved.

Description

Technical Field

[0001] The present application relates to the field related to image data processing technology, and specifically to a detection and recognition method and system based on artificial intelligence. Background Art

[0002] With the continuous development of the automotive industry, the quality and reliability of automotive parts play a vital role in the performance and safety of the entire vehicle. Traditional defect detection methods mainly rely on manual inspection and rule-based image processing technology, which have problems such as low efficiency, insufficient accuracy and high dependence on operators when dealing with complex and diverse defect types.

[0003] Therefore, the existing automotive parts defect detection methods in the prior art have the technical problems of low detection accuracy and efficiency and difficulty in coping with complex and changeable defect types. Summary of the invention

[0004] This application provides an artificial intelligence-based detection and identification method and system to solve the technical problems of the existing automotive parts defect detection methods in the prior art, which have low detection accuracy and efficiency and are difficult to deal with complex and changeable defect types. It significantly improves the accuracy and efficiency of automotive parts defect detection and can more comprehensively identify and classify multiple types of faults.

[0005] The present application provides a detection and identification method based on artificial intelligence, the method comprising: for a target automobile part, calling a fault record and mining a fault feature tree, wherein the fault feature includes a fault area-fault type; based on the fault feature tree, using a multi-level hash analysis based on a pixel array as a detection method, performing sample-driven training, and constructing a defect detection model, wherein the defect detection model comprises a pixel sampling layer, a matrix detection layer, and an evaluation output layer, and a reverse loop is established between the matrix detection layer and the pixel sampling layer; by controlling a front-end sensor device, performing image acquisition on the target automobile part to determine a part image; transmitting the part image back to the defect detection model, performing part defect identification of the pixel array under iterative sampling, and outputting a detection chain; and providing processing guidance for the target automobile part based on the detection chain.

[0006] In the implementation method, the construction of the defect detection model includes: sequentially connecting the pixel sampling layer, the digital array detection layer and the evaluation output layer, and establishing a reverse loop from the digital array detection layer to the pixel sampling layer; setting an initial sampling interval and a sampling adjustment rule, and collaboratively training the pixel sampling layer and the digital array detection layer, wherein the sampling adjustment rule is an interval adjustment ratio.

[0007] In the implementation method, the collaborative training of the pixel sampling layer and the array detection layer includes: the fault feature tree is built in, the fault identification accuracy is used as the convergence condition, and the array detection layer is supervised and trained based on the fault record; the first fault record is extracted to determine the sample original image; based on the initial sampling interval, the sample original image is discretely sampled by pixels to determine the first pixel array; the first pixel array is transferred to the array detection layer to perform fault identification and determine the first pixel fault array; the first pixel fault array is analyzed to locate the fuzzy detection area; for the fuzzy detection area, a second sampling method is determined based on the sampling adjustment rule; the second sampling method is transferred to the pixel sampling layer based on the reverse loop to perform complex sampling to determine the second pixel array of the fuzzy detection area; the second pixel array is transferred to the array detection layer to perform iterative sampling training to complete a training based on the first fault record.

[0008] In the implementation method, based on the initial sampling interval, the sample original image is subjected to discrete pixel sampling to determine a first pixel array, including: obtaining pixel sampling points, wherein the pixel sampling points are characterized as [x, y, z, f(t)], (x, y, z) is the relative distribution position of the sampling points to the target automotive parts, and f(t) is the pixel feature of the sampling points; for the initial sampling interval, N pixel sampling points are determined; based on the relative distribution positions, the N pixel sampling points are arranged in an array to determine the first pixel array.

[0009] In the implementation method, the component defect identification of the pixel array under iterative sampling includes: transmitting the component image back, determining the effective image of the target automobile component by performing background segmentation; based on the pixel sampling layer, performing pixel sampling based on the initial sampling interval on the effective image to determine the target pixel array; transferring the target pixel array to the array detection layer, performing defect determination and positioning and iterative sampling detection, superimposing multiple layers of iterative detection results, and determining the defect detection result; transferring the defect detection result to the evaluation output layer, performing damage degree evaluation and resumption of work evaluation, and determining the detection result.

[0010] In the implementation method, fault records are called and fault feature trees are mined, including: clustering the fault records based on the fault type to determine M groups of fault events; extracting a group of fault events, performing fault feature identification as a root feature layer, refining the root feature layer at multiple levels of granularity to determine a first fault feature tree; traversing the M groups of fault events until the determination of the Mth fault feature tree is completed, integrating the first fault feature tree until the Mth fault feature tree is obtained, and generating the fault feature tree.

[0011] In the implementation method, based on the detection chain, processing guidance of the target automobile parts is carried out, including: identifying the detection chain, wherein the detection chain includes component image-array element digital image-detection result, and the array element digital image includes at least one item; based on the detection chain, determining processing guidance information, wherein the processing guidance includes component resumption guidance and production line processing guidance; based on the processing guidance information, processing guidance is carried out on the target automobile parts.

[0012] The present application also provides an artificial intelligence-based detection and identification system, including: a fault tree construction module, which is used to call fault records and mine fault feature trees for target automobile parts, wherein the fault features include fault area-fault type; a defect detection model construction module, which is used to perform sample-driven training based on the fault feature tree and use multi-level hash analysis based on pixel array as a detection method to construct a defect detection model, wherein the defect detection model includes a pixel sampling layer, a matrix detection layer and an evaluation output layer, and a reverse loop is established between the matrix detection layer and the pixel sampling layer; a component image acquisition module, which is used to perform image acquisition on the target automobile parts by controlling the front-end sensing equipment and determine the component image; a detection chain acquisition module, which is used to transmit the component image back to the defect detection model, perform component defect identification of the pixel array under iterative sampling, and output a detection chain; a processing guidance module, which is used to provide processing guidance for the target automobile parts based on the detection chain.

[0013] The present application proposes a detection and identification method and system based on artificial intelligence, the method comprising: calling fault records and mining fault feature trees for target automotive parts, wherein the fault features include fault area-fault type; based on the fault feature tree, using multi-level hash analysis based on pixel array as the detection method, performing sample-driven training, and building a defect detection model, wherein the defect detection model includes a pixel sampling layer, a matrix detection layer, and an evaluation output layer, and a reverse loop is established between the matrix detection layer and the pixel sampling layer; by controlling the front-end sensor device, the target automotive parts are imaged and the parts image is determined; the parts image is transmitted back to the defect detection model, and the parts defect recognition of the pixel array under iterative sampling is performed, and the detection chain is output; based on the detection chain, the processing guidance of the target automotive parts is performed. The existing automotive parts defect detection method in the prior art solves the technical problem that the detection accuracy and efficiency are low and it is difficult to deal with complex and changeable defect types. The accuracy and efficiency of automotive parts defect detection are significantly improved, and the technical effect of being able to more comprehensively identify and classify multiple types of faults is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the technical solution of the embodiment of the present disclosure, the accompanying drawings of the embodiment of the present disclosure will be briefly introduced below. A flow chart is used in the present application to illustrate the operations performed by the system according to the embodiment of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps can be processed in reverse order or simultaneously as needed. At the same time, other operations can also be added to these processes, or one or more operations can be removed from these processes.

[0015] Figure 1 A flowchart of a detection and identification method based on artificial intelligence is provided in an embodiment of the present application.

[0016] Figure 2 A schematic diagram of the structure of an artificial intelligence-based detection and recognition system provided in an embodiment of the present application.

[0017] Explanation of the reference numerals: fault tree construction module 11 , defect detection model construction module 12 , component image acquisition module 13 , detection chain acquisition module 14 , processing guidance module 15 . DETAILED DESCRIPTION

[0018] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below.

[0019] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.

[0020] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict, and the terms "first\second" involved are merely to distinguish similar objects and do not represent a specific ordering of objects. The terms "including" and "having" and any variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by technicians in the technical field of this application. The terms used herein are for the purpose of describing the embodiments of the present application only.

[0021] The present application provides a detection and recognition method and system based on artificial intelligence, such as Figure 1 As shown, the method includes: for target automobile parts, calling fault records and mining fault feature trees, wherein the fault features include fault area-fault type. Based on the fault feature tree, sample-driven training is performed using multi-level hash analysis based on pixel arrays as a detection method to construct a defect detection model, wherein the defect detection model includes a pixel sampling layer, an array detection layer, and an evaluation output layer, and a reverse loop is established between the array detection layer and the pixel sampling layer.

[0022] In the production process of automobile parts, a certain amount of fault data is usually accumulated. These fault records usually include detailed information of the fault, such as the bit area, type, image data, etc. For the target automobile parts that need to be intelligently detected and identified, the fault records are called and mined to build a fault feature tree. Subsequently, based on the fault feature tree, multi-level hash analysis based on pixel arrays is used as the detection method to perform sample-driven training and build a defect detection model, wherein the defect detection model includes a pixel sampling layer, an array detection layer and an evaluation output layer, and a reverse loop is established between the array detection layer and the pixel sampling layer.

[0023] The method provided by the embodiment of the present application further includes: sequentially connecting the pixel sampling layer, the array detection layer and the evaluation output layer, and establishing a reverse loop from the array detection layer to the pixel sampling layer. Setting an initial sampling interval and a sampling adjustment rule, and collaboratively training the pixel sampling layer and the array detection layer, wherein the sampling adjustment rule is an interval adjustment ratio.

[0024] The basic architecture of the defect detection model is constructed, and the model consists of three main layers: pixel sampling layer, digital array detection layer and evaluation output layer. These layers are connected in sequence to form a feedforward network. In addition, a reverse loop is established between the digital array detection layer and the pixel sampling layer to feed back the detection results to optimize the sampling strategy. Further, the interval distance of the initially selected pixels in the component image is set to obtain the initial sampling interval. At the same time, the sampling adjustment rule is set, mainly the interval adjustment ratio, which is used to dynamically adjust the sampling interval during the iteration process, such as setting the sampling adjustment rule to sample at an interval of 1 / 4 of the initial sampling interval in the fuzzy detection area. The pixel sampling layer and the digital array detection layer are trained collaboratively.

[0025] The method provided in the embodiment of the present application also includes: the fault feature tree is built-in, and the fault identification accuracy is used as the convergence condition, and the array detection layer is supervised and trained based on the fault record. The first fault record is extracted to determine the sample original image. Based on the initial sampling interval, the sample original image is subjected to discrete pixel sampling to determine the first pixel array. The first pixel array is transferred to the array detection layer to perform fault identification and determine the first pixel fault array. The first pixel fault array is analyzed to locate the fuzzy detection area. For the fuzzy detection area, a second sampling method is determined based on the sampling adjustment rule. The second sampling method is transferred to the pixel sampling layer based on the reverse loop, and complex sampling is performed to determine the second pixel array of the fuzzy detection area. The second pixel array is transferred to the array detection layer for iterative sampling training to complete a training based on the first fault record.

[0026] The collaborative training of the pixel sampling layer and the array detection layer includes: integrating the constructed fault feature tree structure into the defect detection model. The fault feature tree organizes fault features in a hierarchical manner to help the model better understand and classify different types of faults. The collected fault records are input into the fault feature tree for supervised training. Supervised training uses labeled data (i.e., records of known fault images and corresponding fault types) to train the model so that it can learn and identify different types of faults. A predetermined fault recognition accuracy rate is set as the convergence condition for model training. When the recognition accuracy rate of the model on the validation set reaches or exceeds this standard, the training is completed and the array detection layer is obtained. The first group of fault records is extracted from the fault record database, and the corresponding fault sample original images are determined based on the records. These images will serve as the basic data of the training model for subsequent pixel sampling and defect recognition. According to the preset initial sampling interval, the sample original image is sampled discretely to form a first pixel array. This process ensures that the model can cover the key areas of the image in the initial stage and improve the comprehensiveness of the detection. The initial sampling interval is the distance between pixels when sampling in the image. Larger intervals help cover a wider area, while smaller intervals help with detailed detection. The first pixel array is an array structure formed by a set of pixel points selected through the initial sampling interval, and is used for subsequent defect identification. The first pixel array is transferred to the array detection layer, and fault identification is performed based on the fault feature tree structure in the array detection layer to determine the first pixel fault array, which is a set of pixel points with faults identified by the array detection layer.

[0027] The first pixel fault array identified is further analyzed to determine which areas have fuzzy or uncertain detection results, and obtain fuzzy detection areas. The fuzzy detection areas are areas where the detection results are not clear enough and further sampling and analysis are required to confirm whether there are defects. These fuzzy detection areas require more detailed sampling to improve the accuracy of detection. For example, there are adjacent pixels, one is faulty and the other is normal. At this time, it is difficult to determine the fault boundary, so it is necessary to sample with a smaller spacing, or migrate the overall position of the initial sampling spacing, and feed back to the sampling layer for resampling analysis until the detection accuracy is met.

[0028] According to the preset sampling adjustment rules, a more detailed second sampling method is determined for the located fuzzy detection area. The second sampling method is transferred to the pixel sampling layer based on the reverse loop, and complex sampling is performed to determine the second pixel array of the fuzzy detection area. The second pixel array is a set of pixel points generated in the fuzzy detection area according to the second sampling method, which is used for further defect identification. The second pixel array is input into the array detection layer for more detailed fault identification. Through iterative sampling and analysis, the detection capability of the model is further optimized, and finally a complete training process based on the first fault record is completed.

[0029] The method provided in the embodiment of the present application also includes: obtaining pixel sampling points, wherein the pixel sampling points are characterized as [x, y, z, f(t)], (x, y, z) is the relative distribution position of the sampling points to the target automotive parts, and f(t) is the pixel feature of the sampling points. For the initial sampling interval, N pixel sampling points are determined. Based on the relative distribution position, the N pixel sampling points are arranged in a digital array to determine the first pixel digital array.

[0030] Based on the initial sampling interval, the sample original image is sampled discretely to determine the first pixel array, including: obtaining pixel sampling points, wherein the pixel sampling points are characterized as [x, y, z, f(t)], (x, y, z) is the relative distribution position of the sampling point to the target automotive parts, and f(t) is the pixel feature of the sampling point. For the initial sampling interval, N pixel sampling points are determined, and further, according to the relative distribution position (x, y, z) of the sampling points, the N pixel sampling points are arranged in an array according to certain rules to form a first pixel array. The first pixel array contains a matrix structure formed by the array arrangement, which contains all sampling point information under the initial sampling interval for subsequent detection and analysis.

[0031] The method provided in the embodiment of the present application also includes: clustering the fault records based on the fault type to determine M groups of fault events. Extract a group of fault events, perform fault feature identification as a root feature layer, perform multi-level granularity refinement on the root feature layer, and determine a first fault feature tree. Traverse the M groups of fault events until the determination of the Mth fault feature tree is completed, integrate the first fault feature tree until the Mth fault feature tree, and generate the fault feature tree.

[0032] Calling fault records and mining fault feature trees includes: analyzing all collected fault records, clustering according to fault types, classifying records with similar fault features (including visual features and fault types) into the same group, thereby simplifying the subsequent feature extraction and analysis process, and determining M groups of fault events, where M is a positive integer greater than 1, and each group of fault events represents a specific type of fault. From the clustered M groups of fault events, extract the fault images of each group one by one, perform detailed fault feature recognition, obtain multiple recognition features, and use them as the root feature layer. Perform multi-level granularity refinement on the root feature layer to determine the first fault feature tree. That is, enrich the features of the root feature layer, such as intercepting, rotating, scaling, translating, etc. It is ensured that the root feature layer after multi-level granularity refinement can cover the display information of different granularities of each feature size, so that no matter what the sampled features are, as long as there is a fault, it can be matched and detected. Finally, traverse the M groups of fault events until the determination of the Mth fault feature tree is completed, integrate the first fault feature tree until the Mth fault feature tree, and generate the fault feature tree.

[0033] By controlling the front-end sensor equipment, the target automobile parts are imaged and the parts images are determined. The parts images are sent back to the defect detection model to perform part defect recognition of the pixel array under iterative sampling and output the detection chain. Based on the detection chain, the processing guidance of the target automobile parts is carried out.

[0034] By controlling the front-end sensing device, the target automobile parts are imaged and the parts image is determined. The front-end sensing device is an industrial camera. The collected parts image is transmitted back to the defect detection model, and the parts defect is identified using the pixel array under iterative sampling, and the detection chain is output. The detection chain records the data stream of the entire detection process, including the results and feedback information of each link such as the parts image, the array element digital image, and the detection results. Finally, based on the detection chain, the processing guidance of the target automobile parts is carried out. The technical problem that the existing automobile parts defect detection method in the prior art has low detection accuracy and efficiency and is difficult to cope with complex and changeable defect types is solved. The accuracy and efficiency of automobile parts defect detection are significantly improved, and the technical effect of more comprehensive identification and classification of multiple types of faults can be achieved.

[0035] The method provided in the embodiment of the present application also includes: transmitting back the component image, and determining the effective image of the target automotive component by performing background segmentation. Based on the pixel sampling layer, pixel sampling based on the initial sampling interval is performed on the effective image to determine the target pixel array. The target pixel array is transferred to the array detection layer to perform defect determination and positioning and iterative sampling detection, and multiple layers of iterative detection results are superimposed to determine the defect detection result. The defect detection result is transferred to the evaluation output layer to perform damage degree evaluation and resumption of work evaluation to determine the detection result.

[0036] The defect recognition of parts and components of pixel arrays under iterative sampling includes: transmitting the collected parts and components image back to the detection system. By using the background segmentation technology, the effective area of ​​the target automobile parts and components is extracted from the image, and the background noise and irrelevant information are removed to ensure the accuracy of subsequent sampling and detection. After the effective image is determined, the pixel sampling layer is used to sample the image according to the preset initial sampling interval to determine the target pixel array. The determined target pixel array is input into the array detection layer, and the pixels are analyzed by using the trained defect detection model. By defect determination and positioning, the corresponding defect category and the position corresponding to the defect category are determined, and the position is determined based on the pixel position, so as to identify the potential defect area. Further, the fuzzy detection area in the potential defect area is located, and the second sampling method is determined by the sampling adjustment rule, and iterative sampling detection is performed, and defect recognition analysis is performed to obtain the recognition result. The multi-layer detection results are superimposed, and the specific location and corresponding category of the defect are finally determined to obtain the defect detection result. The defect detection result is transferred to the evaluation output layer, and the evaluation output layer performs the corresponding damage degree evaluation matching according to the defect detection result, and obtains the corresponding resumption evaluation to determine the detection result. The evaluation output layer records the average damage degree results corresponding to different defect detection results and the corresponding resumption evaluation results of whether to resume production.

[0037] The method provided in the embodiment of the present application further includes: identifying the detection chain, wherein the detection chain includes component image-array element digital image-detection result, and the array element digital image includes at least one item. Based on the detection chain, determining processing guidance information, wherein the processing guidance includes component rework guidance and production line processing guidance. Based on the processing guidance information, providing processing guidance for the target automotive component.

[0038] Based on the detection chain, the processing guidance of the target automotive parts is carried out, including: identifying the detection chain, wherein the detection chain includes the component image-array element digital image-detection result, and the array element digital image includes at least one item. Based on the detection chain, the processing guidance information is determined, wherein the processing guidance includes the component resumption guidance and the production line processing guidance, and the resumption guidance is extracted based on the detection result. The production line processing guidance data is set by professional technicians. Based on the processing guidance information, the target automotive parts are processed and guided.

[0039] In the above, refer to Figure 1 A detection and recognition method based on artificial intelligence according to an embodiment of the present invention is described in detail. Figure 2 An artificial intelligence-based detection and recognition system according to an embodiment of the present invention is described.

[0040] According to an embodiment of the present invention, an artificial intelligence-based detection and identification system solves the technical problems that the existing automotive parts defect detection methods in the prior art have low detection accuracy and efficiency and are difficult to cope with complex and changeable defect types. The accuracy and efficiency of automotive parts defect detection are significantly improved, and the technical effect of being able to more comprehensively identify and classify multiple types of faults is achieved. An artificial intelligence-based detection and identification system includes: a fault tree construction module 11, a defect detection model construction module 12, a component image acquisition module 13, a detection chain acquisition module 14, and a processing guidance module 15.

[0041] The fault tree construction module 11 is used to call the fault record and mine the fault feature tree for the target automobile parts, wherein the fault feature includes the fault area-fault type.

[0042] The defect detection model construction module 12 is used to perform sample-driven training based on the fault feature tree and use multi-level hash analysis based on pixel array as a detection method to construct a defect detection model, wherein the defect detection model includes a pixel sampling layer, an array detection layer and an evaluation output layer, and a reverse loop is established between the array detection layer and the pixel sampling layer.

[0043] The component image acquisition module 13 is used to collect images of the target automobile component by controlling the front-end sensing device to determine the component image.

[0044] The detection chain acquisition module 14 is used to transmit the component image back to the defect detection model, perform component defect recognition on the pixel array under iterative sampling, and output the detection chain.

[0045] The processing guidance module 15 is used to provide processing guidance for the target automobile parts based on the detection chain.

[0046] The specific configuration of the defect detection model construction module 12 will be described in detail below. The defect detection model construction module 12 may further include: the defect detection model is constructed, including: the pixel sampling layer, the array detection layer and the evaluation output layer are sequentially connected, and a reverse loop from the array detection layer to the pixel sampling layer is established. The initial sampling interval and the sampling adjustment rule are set, and the pixel sampling layer and the array detection layer are collaboratively trained, and the sampling adjustment rule is the interval adjustment ratio.

[0047] The specific configuration of the defect detection model construction module 12 will be described in detail below. The defect detection model construction module 12 further includes: the collaborative training of the pixel sampling layer and the array detection layer, including: the built-in fault feature tree, the fault identification accuracy is used as the convergence condition, and the array detection layer is supervised and trained based on the fault record. Extract the first fault record and determine the sample original image. Based on the initial sampling interval, the sample original image is subjected to pixel discrete sampling to determine the first pixel array. The first pixel array is transferred to the array detection layer to perform fault identification and determine the first pixel fault array. The first pixel fault array is analyzed to locate the fuzzy detection area. For the fuzzy detection area, a second sampling method is determined based on the sampling adjustment rule. The second sampling method is transferred to the pixel sampling layer based on the reverse loop to perform complex sampling to determine the second pixel array of the fuzzy detection area. The second pixel array is transferred to the array detection layer to perform iterative sampling training to complete a training based on the first fault record.

[0048] The specific configuration of the defect detection model construction module 12 will be described in detail below. The defect detection model construction module 12 may further include: based on the initial sampling interval, discrete sampling of pixels of the sample original image is performed to determine a first pixel array, including: obtaining pixel sampling points, wherein the pixel sampling points are represented as [x, y, z, f(t)], (x, y, z) is the relative distribution position of the sampling point to the target automotive component, and f(t) is the pixel feature of the sampling point. For the initial sampling interval, N pixel sampling points are determined. Based on the relative distribution position, the N pixel sampling points are arranged in an array to determine the first pixel array.

[0049] The specific configuration of the detection chain acquisition module 14 will be described in detail below. The detection chain acquisition module 14 further includes: the component defect recognition of the pixel array under iterative sampling, including: returning the component image, and determining the effective image of the target automotive component by performing background segmentation. Based on the pixel sampling layer, pixel sampling based on the initial sampling interval is performed on the effective image to determine the target pixel array. The target pixel array is transferred to the array detection layer to perform defect determination and positioning and iterative sampling detection, and multiple layers of iterative detection results are superimposed to determine the defect detection result. The defect detection result is transferred to the evaluation output layer to perform damage degree evaluation and resumption of work evaluation to determine the detection result.

[0050] The specific configuration of the fault tree construction module 11 will be described in detail below. The fault tree construction module 11 further includes: calling fault records and mining fault feature trees, including: clustering the fault records based on the fault type to determine M groups of fault events. Extracting a group of fault events, performing fault feature identification as a root feature layer, performing multi-level granularity refinement on the root feature layer, and determining the first fault feature tree. Traversing the M groups of fault events until the determination of the Mth fault feature tree is completed, integrating the first fault feature tree until the Mth fault feature tree, and generating the fault feature tree.

[0051] The specific configuration of the processing guidance module 15 will be described in detail below. The processing guidance module 15 further includes: based on the detection chain, performing processing guidance for the target automotive parts, including: identifying the detection chain, wherein the detection chain includes a component image-an array element digital image-a detection result, and the array element digital image includes at least one item. Based on the detection chain, determining processing guidance information, wherein the processing guidance includes component resumption guidance and production line processing guidance. Based on the processing guidance information, performing processing guidance on the target automotive parts.

[0052] An artificial intelligence-based detection and identification system provided in an embodiment of the present invention can execute an artificial intelligence-based detection and identification method provided in any embodiment of the present invention, and has functional modules and beneficial effects corresponding to the execution method.

[0053] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules may be used and run on the user terminal and / or server, and the various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention.

[0054] The above specific implementations do not constitute a limitation on the protection scope of this application. It should be understood by those skilled in the art that various modifications, combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of this application should be included in the protection scope of this application.

Claims

1. A detection and recognition method based on artificial intelligence, characterized in that: The method comprises: For the target automotive parts, call the fault records and mine the fault feature tree, where the fault feature includes the fault area-fault type; Based on the fault feature tree, sample-driven training is performed using multi-level hash analysis based on pixel arrays as a detection method to construct a defect detection model, wherein the defect detection model includes a pixel sampling layer, an array detection layer, and an evaluation output layer, and a reverse loop is established between the array detection layer and the pixel sampling layer; By controlling the front-end sensor device, the image of the target automobile part is collected to determine the image of the part; The component image is transmitted back to the defect detection model to perform component defect recognition of the pixel array under iterative sampling, and output a detection chain; Based on the detection chain, processing guidance of the target automobile parts is carried out.

2. The detection and identification method based on artificial intelligence as claimed in claim 1, characterized in that: The step of constructing a defect detection model comprises: The pixel sampling layer, the digital array detection layer and the evaluation output layer are sequentially connected, and a reverse loop from the digital array detection layer to the pixel sampling layer is established; An initial sampling interval and a sampling adjustment rule are set to collaboratively train the pixel sampling layer and the array detection layer, wherein the sampling adjustment rule is an interval adjustment ratio.

3. The detection and identification method based on artificial intelligence as claimed in claim 2, characterized in that: The collaborative training of the pixel sampling layer and the array detection layer includes: The fault feature tree is built in, and the fault identification accuracy is used as a convergence condition, and the digital array detection layer is trained based on the fault record supervision; Extracting the first fault record and determining the sample original image; Based on the initial sampling interval, performing pixel discrete sampling on the sample original image to determine a first pixel array; Transferring the first pixel array to the array detection layer to perform fault identification and determine the first pixel fault array; Analyzing the first pixel fault array to locate the fuzzy detection area; For the blur detection area, determining a second sampling mode based on a sampling adjustment rule; The second sampling mode is transferred to the pixel sampling layer based on the reverse loop to perform complex sampling to determine the second pixel array of the blur detection area; The second pixel array is transferred to the array detection layer to perform iterative sampling training, thereby completing a training based on the first fault record.

4. The detection and identification method based on artificial intelligence as claimed in claim 3, characterized in that: Based on the initial sampling interval, discrete pixel sampling is performed on the sample original image to determine a first pixel array, including: Obtain pixel sampling points, wherein the pixel sampling points are characterized as [x, y, z, f(t)], (x, y, z) is the relative distribution position of the sampling points to the target automotive parts, and f(t) is the pixel feature of the sampling points; Determine N pixel sampling points for the initial sampling interval; Based on the relative distribution positions, the N pixel sampling points are arranged in a digital array to determine the first pixel digital array.

5. The detection and identification method based on artificial intelligence as claimed in claim 2, characterized in that: The component defect recognition of the pixel array under iterative sampling includes: The component image is transmitted back, and a valid image of the target automobile component is determined by performing background segmentation; Based on the pixel sampling layer, performing pixel sampling based on the initial sampling interval on the effective image to determine a target pixel array; The target pixel array is transferred to the array detection layer to perform defect determination and positioning and iterative sampling detection, and multiple layers of iterative detection results are superimposed to determine the defect detection result; The defect detection results are transferred to the evaluation output layer to perform damage degree evaluation and resumption of work evaluation to determine the detection results.

6. The detection and identification method based on artificial intelligence as claimed in claim 1, characterized in that: Call fault records and mine fault feature trees, including: Performing clustering processing on the fault records based on the fault type to determine M groups of fault events; Extracting a group of fault events, performing fault feature identification as a root feature layer, performing multi-level granularity refinement on the root feature layer, and determining a first fault feature tree; The M groups of fault events are traversed until the determination of the Mth fault feature tree is completed, and the first fault feature trees are integrated until the Mth fault feature tree, to generate the fault feature tree.

7. The detection and identification method based on artificial intelligence as claimed in claim 1, characterized in that: Based on the detection chain, the processing guidance of the target automobile parts is carried out, including: Identify the detection chain, wherein the detection chain includes component image-array element digital image-detection result, and the array element digital image includes at least one item; Based on the detection chain, determine processing guidance information, where the processing guidance includes component resumption guidance and production line processing guidance; Based on the processing guidance information, processing guidance is provided for the target automobile component.

8. A detection and recognition system based on artificial intelligence, characterized in that: The system comprises: The fault tree construction module is used to call the fault records and mine the fault feature tree for the target automobile parts, where the fault feature includes the fault area-fault type; A defect detection model construction module is used to perform sample-driven training based on the fault feature tree and multi-level hash analysis based on pixel array as a detection method to construct a defect detection model, wherein the defect detection model includes a pixel sampling layer, a matrix detection layer and an evaluation output layer, and a reverse loop is established between the matrix detection layer and the pixel sampling layer; A component image acquisition module is used to collect images of the target automobile component by controlling the front-end sensing device to determine the component image; A detection chain acquisition module is used to transmit the component image back to the defect detection model, perform component defect recognition of the pixel array under iterative sampling, and output a detection chain; A processing guidance module is used to provide processing guidance for the target automobile parts based on the detection chain.

Images