Vehicle disassembly part tracing method and device, terminal and storage medium

Abstract

The invention discloses a vehicle disassembled part traceability method and device, a terminal and a storage medium, and the method comprises the steps: carrying out the omnibearing image collection of a disassembled part, obtaining an appearance image, carrying out the material component analysis, and obtaining an appearance image and a material analysis result; judging whether the disassembled part has an identification code or not according to the appearance image, searching source information of the current disassembled part from a preset part database when the identification code exists on the disassembled part, and outputting a traceability result of the disassembled part according to the source information; when the source information does not exist or the identification code does not exist on the disassembled part, performing feature extraction on the appearance image and the material analysis result to obtain a comprehensive feature vector; and comparing the comprehensive feature vector with a part database, and determining and outputting a traceability result of the disassembled part according to a comparison result. According to the method, the feature information such as the appearance, the size, the material composition and the identification code of the disassembled part is comprehensively analyzed, so that the source of the disassembled part is quickly and accurately traced.

Description

Technical Field

[0001] The present application relates to the technical field of vehicle dismantling management, and in particular to a vehicle dismantling parts traceability method, device, terminal and storage medium. Background Art

[0002] At present, the dismantling of scrapped vehicles mainly relies on manual operations. During the dismantling process, it is often difficult to effectively mark and record the dismantled parts, such as engines and gearboxes. This traditional manual recording method is not only inefficient, but also easily leads to information loss or confusion during the dismantling process, which brings great trouble to the subsequent parts traceability and reuse. The existing solution is mainly to mark the dismantled parts through manual recording during the dismantling process. However, this method is not only labor-intensive, but also prone to errors, and cannot guarantee the accuracy and completeness of the recorded information. The manual dismantling and recording method has obvious efficiency and accuracy problems, and cannot meet the high requirements of modern scrapped vehicle dismantling companies for parts traceability. At the same time, this method also increases the operating costs of enterprises and restricts the sustainable development of the scrapped vehicle dismantling industry.

[0003] In summary, it is particularly important to provide a method that can accurately and automatically record disassembled parts information and achieve efficient traceability. Summary of the invention

[0004] The present application provides a vehicle dismantled parts traceability method, device, terminal and storage medium, which realizes rapid and accurate traceability of the sources of dismantled parts by comprehensively analyzing characteristic information such as the shape, size, material composition and identification code of the dismantled parts.

[0005] In the first aspect, the present application provides a vehicle dismantling parts traceability method, which adopts the following technical solutions: A vehicle dismantling parts tracing method comprises the following steps: Performing omnidirectional image acquisition on the disassembled parts to obtain appearance images, performing material composition analysis on the disassembled parts to obtain material analysis results, and summarizing the appearance images and the material analysis results to obtain acquisition information; Determine whether the disassembled part has an identification code according to the appearance image; if the disassembled part has an identification code, search for source information of the current disassembled part from a preset parts database according to the identification code; if the source information exists, output a traceability result of the current disassembled part according to the source information; When the source information does not exist or the identification code does not exist on the disassembled part, feature extraction is performed on the collected information to obtain a comprehensive feature vector; The comprehensive feature vector is compared with the parts database to obtain a comparison result, and the traceability result of the current disassembled part is determined and output based on the comparison result.

[0006] By adopting the above-mentioned technical means and combining multiple features including appearance, material, and identification code, the accuracy of traceability is greatly improved, which can significantly reduce the risk of information loss, confusion or error, and ensure the reliability of traceability results; when there is an identification code on the part, the part database is directly queried through the identification code to achieve fast and accurate traceability; when the identification code is missing or unrecognizable, the appearance features, size features and material features of the part are extracted and compared with the parts database, so as to accurately find the source of the part, thereby ensuring the applicability and flexibility of this method and further improving the accuracy of traceability.

[0007] Preferably, judging whether the disassembled part has an identification code based on the appearance image, and when the disassembled part has an identification code, searching for source information of the current disassembled part from a preset parts database based on the identification code, and when the source information exists, outputting the traceability result of the current disassembled part based on the source information, specifically comprises the following steps: Preprocessing the appearance image, detecting the preprocessed appearance image, and determining whether an identification code exists for the corresponding disassembled part; When there is an identification code in the appearance image of the disassembled part, scanning and identifying the identification code to obtain production information of the disassembled part, the production information including a production batch and a serial number; sequentially querying matching source information in a preset parts database according to the production batch and the serial number, wherein the parts database contains characteristic information of all automobile parts of known manufacturers; When the matching source information is found, the traceability result of the current disassembled part is directly outputted according to the source information.

[0008] By adopting the above-mentioned technical means, the appearance image is preprocessed and detected, and it is quickly determined whether there is an identification code on the disassembled parts, avoiding the tedious process of manual input or search for information in the traditional traceability method, and greatly improving the traceability efficiency; when there is an identification code on the part, the parts database is directly queried through the identification code, further realizing fast and accurate traceability.

[0009] Preferably, when the source information does not exist or the identification code does not exist on the disassembled parts, feature extraction is performed on the collected information to obtain a comprehensive feature vector, which specifically includes the following steps: Feature extraction is performed on the appearance image and the material analysis result, specifically, Obtaining appearance features of the disassembled parts according to the appearance image, wherein the appearance features include shape features and size features; and generating material features according to the material analysis results; The shape feature, the size feature and the material feature are fused to obtain a comprehensive feature vector.

[0010] By adopting the above-mentioned technical means, when the identification code is missing or cannot be recognized, the source of the parts can still be accurately found by extracting the appearance features, size features and material features of the disassembled parts and comparing them with the parts database, which has higher applicability and flexibility; and the fused features contain a variety of information such as shape, size, material, etc., which can realize the subsequent comprehensive description and accurate identification of the disassembled parts, further improving the accuracy of traceability.

[0011] Preferably, before comparing the comprehensive feature vector with the parts database, the following steps are also included: The comprehensive feature vector is standardized, specifically, a data set of the comprehensive feature vector is constructed according to the shape feature, the size feature and the material feature, that is, X=[X Shape ,X Size ,X Material ]; Among them, X Shape is the value of the shape feature, X Size For the size feature, X Material is the numerical value of the material characteristic; The shape feature and the material feature are standardized, that is, Among them, μ Shape is the mean of the shape characteristics, σ Shape is the standard deviation of the shape feature, μ Material is the mean value of the material characteristic, σ Material is the standard deviation of the material characteristic; The size characteristics are normalized, i.e., Where i = 1, 2, ..., n; n is the number of the size features, μ Size,i is the mean value of each of the size characteristics, σ Size,i is the standard deviation of each of the said size characteristics; The standardized comprehensive feature vector is obtained, that is, X′=[X′ Shape ,X′ Material ,X′ Size,i].

[0012] By adopting the above technical means, the comprehensive feature vector is standardized so that the feature vectors of different automobile parts can be directly compared, avoiding the influence of the comparison on the accuracy of the result.

[0013] Preferably, the step of comparing the comprehensive feature vector with the parts database and obtaining a comparison result specifically comprises the following steps: The shape feature, the size feature and the material feature in the comprehensive feature vector are sequentially traversed, each feature having a corresponding weight coefficient, and a comparison algorithm for each feature is selected from a preset algorithm selection list according to each feature and the corresponding weight coefficient; The shape feature corresponds to a shape matching algorithm, the size feature corresponds to a distance measurement algorithm, and the material feature corresponds to a classification algorithm; The priority of the comparison algorithm is determined according to the weight coefficient of each feature in the comprehensive feature vector, thereby determining the currently selected comparison algorithm.

[0014] By adopting the above-mentioned technical means, we can ensure that the importance of different features can be fully considered when selecting algorithms, so as to make more reasonable and accurate decisions, which will help improve the accuracy and reliability of traceability, and also help optimize computing resources and time costs.

[0015] Preferably, the method further comprises the following steps: Extracting a standard shape feature of each automobile part from the parts database, calculating a first similarity between the shape feature and the standard shape feature according to the comparison algorithm, screening the automobile parts whose first similarity is greater than a preset first similarity threshold from the parts database, and summarizing them into a first candidate parts list; Extracting the standard size feature of each first candidate part from the parts database, calculating the second similarity between the size feature and the standard size feature according to the comparison algorithm, screening out the automobile parts whose second similarity is greater than a preset second similarity threshold from the first candidate parts, and summarizing them into a second candidate parts list; The standard material feature of each second candidate part is extracted from the parts database, the third similarity between the material feature and the standard material feature is calculated according to the comparison algorithm, and the automotive parts whose third similarity is greater than a preset third similarity threshold are screened from the second candidate parts, and they are summarized into a third candidate parts table.

[0016] By adopting the above-mentioned technical means, through a step-by-step screening method, first based on shape features, then size features, and finally material features, it is possible to efficiently narrow the selection range from a huge parts database and quickly locate the candidate parts that best meet the requirements, avoiding blind searches, improving matching efficiency, and reducing unnecessary time and resource consumption; and the similarity threshold can be adjusted according to actual needs to achieve higher screening accuracy.

[0017] Preferably, determining and outputting the traceability result of the current disassembled part according to the comparison result specifically comprises the following steps: Arrange the third candidate parts list in descending order to obtain a traceability parts list of the current disassembled parts; According to a preset candidate threshold, the automobile parts having a quantity equal to the candidate threshold are selected from top to bottom in the traceability parts list, and the source information corresponding to the selected automobile parts is searched from the parts database; The selected automobile parts and the corresponding source information are summarized into the traceability results of the current disassembled parts and output.

[0018] By adopting the above-mentioned technical means, the third candidate parts list is arranged in descending order of similarity, and the resulting traceability parts list clearly displays the candidate parts that best match the disassembled parts, making the traceability results more intuitive and easy to understand; and based on the preset candidate threshold, a certain number of automotive parts are selected from top to bottom, ensuring that the selected parts meet the similarity requirements and can be controlled within a reasonable quantity range, thereby improving selection efficiency.

[0019] In the second aspect, the present application provides a vehicle dismantling parts tracing device, which adopts the following technical solution: A vehicle dismantling parts tracing device includes the following modules: An information acquisition module is used to perform omnidirectional image acquisition on the disassembled parts to obtain an appearance image, perform material composition analysis on the disassembled parts to obtain a material analysis result, and summarize the appearance image and the material analysis result to obtain acquisition information; an identification code traceability module is used to determine whether the disassembled parts have an identification code based on the appearance image, and when the disassembled parts have an identification code, search for source information of the current disassembled parts from a preset parts database based on the identification code, and when the source information exists, output the traceability result of the current disassembled parts based on the source information; A feature extraction module, used for extracting features from the collected information to obtain a comprehensive feature vector when the source information does not exist or the identification code does not exist on the disassembled parts; The comparison and tracing module is used to compare the comprehensive feature vector with the parts database and obtain a comparison result, and determine and output the tracing result of the current disassembled part based on the comparison result.

[0020] By adopting the above-mentioned technical means, a system for tracing the dismantled parts of scrapped vehicles has been established, providing a more transparent and efficient parts management method for the scrapped vehicle dismantling industry, which is conducive to the standardization and standardization of the industry.

[0021] In a third aspect, the present application provides a smart terminal, which adopts the following technical solution: A smart terminal includes a memory and a processor, wherein the memory stores at least one instruction, at least one program, code set or instruction set, and the at least one instruction, at least one program, code set or instruction set is loaded and executed by the processor to implement the vehicle dismantled parts traceability method as described above.

[0022] In a fourth aspect, the present application provides a computer-readable storage medium, which adopts the following technical solution: A computer-readable storage medium, wherein at least one instruction, at least one program, code set or instruction set is stored in the readable storage medium, and the at least one instruction, at least one program, code set or instruction set is loaded and executed by a processor to implement the vehicle dismantled parts traceability method as described above.

[0023] In summary, this application has at least the following beneficial effects: 1. This application comprehensively considers various information such as the shape, size, material composition and identification code of the disassembled parts through multi-feature fusion technology, which greatly improves the accuracy of traceability, can significantly reduce the risk of information loss, confusion or error, and ensure the reliability of traceability results.

[0024] 2. This application uses automated image acquisition and material analysis to quickly obtain the appearance and material information of disassembled parts, significantly improving the efficiency of traceability operations; when there is an identification code on the part or the identification code cannot be identified, the source of the part can be accurately found, ensuring the applicability and flexibility of this method, and further improving the accuracy of traceability.

[0025] 3. The entire traceability process of this application is highly automated, from image acquisition, material analysis to feature extraction and database comparison, which reduces manual intervention, improves traceability efficiency, and reduces the risk of human error. It can handle a variety of traceability scenarios, which is crucial to the subsequent reuse management of scrapped parts and promotes sustainable development. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 This is a flow chart of a vehicle dismantling parts tracing method according to an embodiment of the present application; Figure 2 It is an architectural diagram of a vehicle dismantling parts traceability device according to an embodiment of the present application. DETAILED DESCRIPTION

[0027] The present application provides a vehicle dismantling parts traceability method, device, terminal and storage medium. In order to make the purpose, technical solution and advantages of the present application clearer, the implementation method of the present application will be further described in detail below.

[0028] The following is a further detailed description of an embodiment of a vehicle dismantling parts traceability method of the present application in conjunction with the drawings of the specification.

[0029] A vehicle dismantling parts tracing method of the present application, such as Figure 1 As shown, the following steps are included: S1. Set up a high-resolution camera or 3D scanner to collect all-round images of the disassembled parts to obtain appearance images. There are multiple appearance images, and each image can identify the shape contour and size information of the disassembled parts.

[0030] Use a spectrum analyzer or X-ray fluorescence spectrometer to perform material composition analysis on the disassembled parts to obtain material analysis results and determine the main components of the disassembled parts and possible alloy types.

[0031] The appearance images and material analysis results are summarized to obtain the collected information.

[0032] S2. Determine whether there is an identification code on the disassembled part based on the appearance image. When there is an identification code on the disassembled part, search for source information of the current disassembled part from a preset parts database based on the identification code. When source information exists, output the traceability result of the current disassembled part based on the source information, which specifically includes the following steps.

[0033] S21, preprocessing the appearance image, the preprocessing includes denoising, contrast enhancement and edge detection to improve the accuracy of subsequent feature extraction, which will not be described in detail here.

[0034] S22, detecting the pre-processed appearance image to determine whether the corresponding disassembled part has an identification code, where the identification code includes an electronic tag or a QR code.

[0035] Use edge detection algorithm to identify edges in the image, determine the approximate position and shape of the identification code, and find matching shapes in the edge detection results based on the known shape of the identification code. Based on the shape matching results, determine whether there is an area in the image that matches the identification code features. If the match is successful, it exists, otherwise it does not exist.

[0036] S23. When there is an identification code in the appearance image of the disassembled part, the identification code is scanned and identified by an RFID reader or a two-dimensional code scanner to obtain production information of the disassembled part, where the production information includes a production batch and a serial number.

[0037] S24. Before searching, it is necessary to first construct a parts database containing various characteristic information of all automobile parts of known manufacturers. This parts database contains the following information: Manufacturer information: including the manufacturer's name, address and contact information.

[0038] Automotive parts feature information: including the parts’ outline, dimensions, material composition, and identification codes. This information has been standardized to ensure that data from different sources can be compared.

[0039] Other relevant information: such as production date, production batch, serial number and quality certification, can help further confirm the source and quality of parts.

[0040] S25. Search for matching source information in a preset parts database according to the production batch and serial number in turn, first search according to the production batch, and then filter the searched results by serial number.

[0041] Source information includes manufacturer information, standard feature information of parts, and other relevant information including production date, serial number, and quality certification.

[0042] S25. When matching source information is found, the traceability result of the current disassembled parts is directly output based on the source information.

[0043] S3. When there is no source information or no identification code on the disassembled parts, feature extraction is performed on the collected information to obtain a comprehensive feature vector, which specifically includes the following steps.

[0044] S31, extracting features from the appearance image and material analysis results, specifically, The appearance features of the disassembled parts are obtained based on the appearance images, and the appearance features include shape features and size features.

[0045] The shape features are extracted by a preset image processing technology, and this embodiment uses an edge detection algorithm. The shape features may include the contour, edge or key point positions of the disassembled parts and their mutual relationships, which are helpful for identifying disassembled parts of different shapes.

[0046] The dimensional features are extracted through the dimensional measurement function in the image processing technology. The dimensional features may include the length, width, height and diameter of the disassembled parts. These features are crucial for distinguishing disassembled parts of different sizes.

[0047] Generate material characteristics based on material analysis results.

[0048] Material characteristics are extracted through material identification technology, including spectral analysis and chemical analysis. In this embodiment, X-ray fluorescence spectroscopy (XRF) non-destructive testing technology is used to identify the composition of the material, which will not damage the disassembled parts and the disassembled parts can be subsequently recycled.

[0049] Material characteristics can include physical and chemical properties such as composition, structure, hardness or elastic modulus of the disassembled part.

[0050] S32, fusing the shape features, size features and material features to obtain a comprehensive feature vector.

[0051] The feature fusion method used in this embodiment is a feature weighting method, which assigns different weight coefficients to different features according to the importance of shape features, size features and material features, and the weight coefficients are adjusted according to actual needs.

[0052] S4. Standardize the comprehensive feature vector.

[0053] The feature vectors of different automobile parts may have different dimensions and value ranges. Direct comparison may lead to inaccurate results. Therefore, before comparing the comprehensive feature vector with the parts database, the comprehensive feature vector needs to be standardized, which includes the following steps.

[0054] S41. Construct a data set of comprehensive feature vectors based on shape features, size features and material features, that is, X=[X Shape ,X Size ,X Material ]; Among them, X Shape is the value of the shape feature, X Size is the value of the size feature, X Material is the numerical value of the material characteristic.

[0055] S42. Standardize the shape characteristics and material characteristics, that is, Among them, μ Shape is the mean of the shape feature, σ Shape is the standard deviation of the shape feature, μ Material is the mean of the material characteristics, σ Material is the standard deviation of the material characteristic.

[0056] S43, standardize the size characteristics, that is, Where i = 1, 2, ..., n; n is the number of size features, μ Size,i is the mean of each size feature, σ Size,i is the standard deviation of each size feature.

[0057] S44, get the standardized comprehensive feature vector, that is, X′=[X′ Shape ,X′ Material ,X′ Size,i ].

[0058] Among them, X′ Shape is the normalized shape feature, X′ Material Normalized material characteristics, X′ Size,i is the standardized size feature.

[0059] The following is a simplified example, assuming that there is a comprehensive feature vector X of the disassembled parts as follows: Shape characteristics (X Shape ): contour complexity is 60; Size feature (X Size ): length is 120, width is 60, height is 25; Material characteristics (X Material ): The main material component index is 80; The mean μ of the shape feature Shape =50, standard deviation σ Shape =15, the mean value of the size characteristic μ Size,i =[100,50,20], standard deviation of size characteristics σ Size,i =[20,10,5], Mean μ of material characteristics Material =70, standard deviation of material characteristics σ Material =10.

[0060] The above values ​​need to be obtained through a large amount of data statistics, and are directly given in this embodiment for the purpose of example.

[0061] Shape characteristics (X′ Shape ): (60-50) / 15=0.67 (keep two decimal places); Size feature (X′ Size,i ): [(120-100) / 20, (60-50) / 10, (25-20) / 5] = [1.00, 1.00, 1.00]; Material characteristics (X′ Material ): (80-70) / 10=1.00; Therefore, the standardized comprehensive feature vector X′=[0.67,[1.00,1.00,1.00],1.00].

[0062] S5. Compare the comprehensive feature vector with the parts database and obtain the comparison result, and determine and output the traceability result of the current disassembled parts based on the comparison result, which specifically includes the following steps.

[0063] S51. The shape features, size features and material features in the comprehensive feature vector are traversed in sequence, each feature having a corresponding weight coefficient, which has been allocated when the comprehensive feature vector is generated.

[0064] A comparison algorithm for each feature is selected from a preset algorithm selection list based on each feature and the corresponding weight coefficient.

[0065] The shape feature corresponds to a shape matching algorithm, and the shape matching algorithm of this embodiment uses Hausdorff distance.

[0066] The size feature corresponds to a distance measurement algorithm, and the distance measurement algorithm of this embodiment uses the Euclidean distance.

[0067] The material characteristics correspond to the classification algorithm, and the classification algorithm of this embodiment uses a support vector machine (SVM).

[0068] S52. Combining the weight coefficient of each feature in the comprehensive feature vector, determining the priority of the comparison algorithm according to the weight coefficient, thereby determining the currently selected comparison algorithm.

[0069] The higher the weight coefficient of a feature, the more important the feature is in the traceability process. When selecting a comparison algorithm, the decision logic is adjusted according to the weight coefficient so that the algorithm corresponding to the feature with a higher weight coefficient occupies a larger proportion in the final selection.

[0070] For example, in a specific feasible implementation, it is determined that shape features are very critical in the traceability process and can significantly distinguish different parts, so a higher weight coefficient is assigned to the shape features. When selecting a comparison algorithm, priority is given to algorithms that can handle shape features well, such as shape matching algorithms. In addition, if a comparison algorithm performs particularly well on shape features, then even if it performs averagely on other features, it will be selected because of the high weight of the shape feature.

[0071] S53, calculating the similarity result between the comprehensive feature vector and the parts database according to a preset comparison algorithm, specifically comprising the following steps: S531. Extract standard shape features of each automobile part from the parts database, and calculate a first similarity between the shape features and the standard shape features according to a comparison algorithm.

[0072] The comparison algorithm used here is the comparison algorithm selected through the above steps.

[0073] S532: Filter out automobile parts whose first similarity is greater than a preset first similarity threshold from the parts database, and summarize them into a first candidate parts list.

[0074] The first similarity threshold is adjusted according to actual needs during specific implementation. When there is a requirement for the accuracy of the screening result or the screening speed, the first similarity threshold is increased.

[0075] S533. Extract the standard size feature of each first candidate part from the parts database, and calculate the second similarity between the size feature and the standard size feature according to the comparison algorithm.

[0076] S534: Filter out automobile parts whose second similarity is greater than a preset second similarity threshold from the first candidate parts, and summarize them into a second candidate parts list.

[0077] The second similarity threshold is adjusted according to actual needs during specific implementation.

[0078] S535. Extract the standard material feature of each second candidate part from the parts database, and calculate the third similarity between the material feature and the standard material feature according to the comparison algorithm.

[0079] S536: Filter out automobile parts whose third similarity is greater than a preset third similarity threshold from the second candidate parts, and summarize them into a third candidate parts list.

[0080] The third similarity threshold is adjusted according to actual needs during specific implementation.

[0081] In order to control the accuracy of the traceability results, three similarity thresholds are set. Only when the similarity between the comprehensive feature vector and the feature vector in the parts database is higher than the corresponding similarity threshold, the match is considered successful. Through layer-by-layer screening, the auto parts that meet the requirements are gradually screened out, which can save computing resources and improve work speed.

[0082] S54. Arrange the third candidate parts list in descending order to obtain a traceability parts list of the current disassembled parts.

[0083] According to the preset candidate threshold, automobile parts with a number equal to the candidate threshold are selected from the traceability parts list from top to bottom, and the source information of the selected automobile parts is searched from the parts database.

[0084] The candidate threshold is adjusted according to the needs in the actual implementation, but it should not be greater than 10 to avoid the situation where the candidate threshold is greater than the number of automobile parts in the traceability parts list.

[0085] S55. The selected automobile parts and corresponding source information are summarized into the traceability results of the current disassembled parts and output.

[0086] The traceability results include the manufacturer information of the selected automotive parts, detailed part information and comparison results.

[0087] The manufacturer information is the name and address of the manufacturer of the successfully matched automobile parts; Parts details include production date, production batch, serial number and quality certification; The comparison result includes the calculation result of the similarity and the position or number of the successfully matched automobile part in the parts database.

[0088] Based on the same inventive concept as above, the present application also discloses a vehicle dismantling parts tracing device, the structure of which is as follows: Figure 2 As shown, the device includes the following modules: An information collection module is used to collect all-round images of the disassembled parts to obtain appearance images, perform material composition analysis on the disassembled parts to obtain material analysis results, and summarize the appearance images and material analysis results to obtain collected information; The identification code traceability module is used to determine whether the disassembled part has an identification code based on the appearance image. When the disassembled part has an identification code, the source information of the current disassembled part is searched from the preset parts database based on the identification code. When the source information exists, the traceability result of the current disassembled part is output based on the source information. A feature extraction module is used to extract features from the collected information to obtain a comprehensive feature vector when there is no source information or no identification code on the disassembled parts; The comparison and traceability module is used to compare the comprehensive feature vector with the parts database and obtain the comparison result, and determine and output the traceability result of the current disassembled part based on the comparison result.

[0089] In a specific implementation scheme, the identification code traceability module includes the following units: The first identification code traceability unit is used to preprocess the appearance image, detect the preprocessed appearance image, and determine whether the corresponding disassembled part has an identification code; when the identification code exists in the appearance image of the disassembled part, scan and identify the identification code to obtain the production information of the disassembled part, and the production information includes the production batch and serial number; The second identification code traceability unit is used to query matching source information in a preset parts database according to the production batch and serial number in turn. The parts database contains characteristic information of all automobile parts of known manufacturers; when matching source information is queried, the traceability result of the current disassembled part is directly output based on the source information.

[0090] In a specific implementation scheme, the feature extraction module includes the following units: The first feature extraction unit is used to extract features from the appearance image and the material analysis result, specifically, Obtaining appearance features of the disassembled parts based on the appearance images, the appearance features include shape features and size features; Generate material characteristics based on material analysis results.

[0091] The second feature extraction unit is used to fuse the shape feature, the size feature and the material feature to obtain a comprehensive feature vector.

[0092] In a specific implementation scheme, the comparison and tracing module includes the following units: The first comparison and tracing unit is used to sequentially traverse the shape features, size features and material features in the comprehensive feature vector, each feature having a corresponding weight coefficient, and select a comparison algorithm for each feature in a preset algorithm selection list according to each feature and the corresponding weight coefficient; Shape features correspond to shape matching algorithms, size features correspond to distance measurement algorithms, and material features correspond to classification algorithms; Combined with the weight coefficient of each feature in the comprehensive feature vector, the priority of the comparison algorithm is determined according to the weight coefficient, thereby determining the currently selected comparison algorithm.

[0093] The second comparison and tracing unit is used to extract the standard shape feature of each automobile part from the parts database, calculate the first similarity between the shape feature and the standard shape feature according to the comparison algorithm, screen out automobile parts whose first similarity is greater than a preset first similarity threshold from the parts database, and summarize them into a first candidate parts list; Extracting the standard size feature of each first candidate part from the parts database, calculating the second similarity between the size feature and the standard size feature according to the comparison algorithm, screening out automobile parts whose second similarity is greater than a preset second similarity threshold from the first candidate parts, and summarizing them into a second candidate parts list; The standard material features of each second candidate part are extracted from the parts database, the third similarity between the material features and the standard material features is calculated according to the comparison algorithm, and the automotive parts whose third similarity is greater than a preset third similarity threshold are screened from the second candidate parts, and they are summarized into a third candidate parts table.

[0094] A third comparison and tracing unit is used to sort the third candidate parts list in descending order to obtain a traceability parts list of the currently disassembled parts; According to the preset candidate threshold, select the automobile parts with the number equal to the candidate threshold from the top to the bottom in the traceability parts list, and search the source information corresponding to the selected automobile parts from the parts database; The selected automobile parts and corresponding source information are summarized into the traceability results of the current disassembled parts and output.

[0095] In a specific feasible implementation scheme, a vehicle dismantling parts tracing device further includes the following modules: The vector normalization module is used to normalize the comprehensive feature vector, specifically, A data set of comprehensive feature vectors is constructed based on shape features, size features, and material features, that is, X=[X Shape ,X Size ,X Material ]; Among them, X Shape is the value of the shape feature, X Size is the size feature, X Material is the numerical value of the material characteristic; Normalize the shape and material characteristics, i.e. Among them, μ Shape is the mean of the shape feature, σ Shape is the standard deviation of the shape feature, μ Material is the mean of the material characteristics, σ Material is the standard deviation of the material characteristic; Standardize the size characteristics, that is, Where i = 1, 2, ..., n; n is the number of size features, μ Size,i is the mean of each size feature, σ Size,i is the standard deviation of each size feature; The standardized comprehensive feature vector is obtained, that is, X′=[X′ Shape ,X′ Material ,X′ Size,i ].

[0096] From the above functional introduction, it can be seen that the vehicle dismantling parts traceability device in this application has built a tracking and tracing system for scrapped automobile parts, providing a more transparent and efficient parts management method for the scrapped automobile dismantling industry, which is conducive to the standardization and standardization of the industry and promotes the sustained and rapid development of the economy and society.

[0097] Based on the same inventive concept mentioned above, an embodiment of the present application also discloses a computer-readable storage medium, which stores at least one instruction, at least one program, code set or instruction set. The at least one instruction, at least one program, code set or instruction set can be loaded and executed by a processor to implement the vehicle dismantled parts traceability method provided in the above method embodiment.

[0098] Also based on the same inventive concept mentioned above, an embodiment of the present application also discloses a computer-readable storage medium, in which at least one instruction, at least one program, code set or instruction set is stored, and at least one instruction, at least one program, code set or instruction set is loaded and executed by a processor to implement the vehicle dismantled parts traceability method as mentioned above.

[0099] A person skilled in the art can understand that all or part of the steps of implementing the above embodiments can be completed by hardware, or can be completed by a program to instruct the relevant hardware, and the program can be stored in a computer-readable storage medium, and the computer-readable storage medium includes, for example: a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.

[0100] The above are only optional embodiments of the present application and are not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.

Claims

1. A vehicle dismantling parts tracing method, characterized in that: The steps include: Performing omnidirectional image acquisition on the disassembled parts to obtain appearance images, performing material composition analysis on the disassembled parts to obtain material analysis results, and summarizing the appearance images and the material analysis results to obtain acquisition information; Determine whether the disassembled part has an identification code according to the appearance image; if the disassembled part has an identification code, search for source information of the current disassembled part from a preset parts database according to the identification code; if the source information exists, output a traceability result of the current disassembled part according to the source information; When the source information does not exist or the identification code does not exist on the disassembled part, feature extraction is performed on the collected information to obtain a comprehensive feature vector; The comprehensive feature vector is compared with the parts database to obtain a comparison result, and the traceability result of the current disassembled part is determined and output based on the comparison result.

2. The vehicle dismantling parts tracing method according to claim 1 is characterized in that: The method of judging whether the disassembled part has an identification code based on the appearance image, and when the disassembled part has an identification code, searching for source information of the current disassembled part from a preset parts database based on the identification code, and when the source information exists, outputting a traceability result of the current disassembled part based on the source information, specifically includes the following steps: Preprocessing the appearance image, detecting the preprocessed appearance image, and determining whether an identification code exists for the corresponding disassembled part; When there is an identification code in the appearance image of the disassembled part, scanning and identifying the identification code to obtain production information of the disassembled part, the production information including a production batch and a serial number; sequentially querying matching source information in a preset parts database according to the production batch and the serial number, wherein the parts database contains characteristic information of all automobile parts of known manufacturers; When the matching source information is found, the traceability result of the current disassembled part is directly outputted according to the source information.

3. The vehicle dismantling parts tracing method according to claim 1 is characterized in that: When the source information does not exist or the identification code does not exist on the disassembled parts, feature extraction is performed on the collected information to obtain a comprehensive feature vector, which specifically includes the following steps: Feature extraction is performed on the appearance image and the material analysis result, specifically, Obtaining appearance features of the disassembled parts according to the appearance image, wherein the appearance features include shape features and size features; generating material characteristics according to the material analysis results; The shape feature, the size feature and the material feature are fused to obtain a comprehensive feature vector.

4. The vehicle dismantling parts tracing method according to claim 3 is characterized in that: Before comparing the comprehensive feature vector with the parts database, the method further includes the following steps: The comprehensive feature vector is standardized, specifically, The data set of the comprehensive feature vector is constructed according to the shape feature, the size feature and the material feature, that is, X=[X Shape ,X Size ,X Material ]; Among them, X Shape is the value of the shape feature, X Size For the size feature, X Material is the numerical value of the material characteristic; The shape feature and the material feature are standardized, that is, Among them, μ Shape is the mean of the shape feature, σ Shape is the standard deviation of the shape feature, μ Material is the mean value of the material characteristic, σ Material is the standard deviation of the material characteristic; The size characteristics are normalized, i.e., Where i = 1, 2, ..., n; n is the number of the size features, μ Size,i is the mean value of each of the size characteristics, σ Size,i is the standard deviation of each of the said size characteristics; The standardized comprehensive feature vector is obtained, that is, X′=[X′ Shape ,X′ Material ,X′ Size,i ]。 5. The vehicle dismantling parts tracing method according to claim 3 is characterized in that: The step of comparing the comprehensive feature vector with the parts database and obtaining a comparison result specifically includes the following steps: The shape feature, the size feature and the material feature in the comprehensive feature vector are sequentially traversed, each feature having a corresponding weight coefficient, and a comparison algorithm for each feature is selected from a preset algorithm selection list according to each feature and the corresponding weight coefficient; The shape feature corresponds to a shape matching algorithm, the size feature corresponds to a distance measurement algorithm, and the material feature corresponds to a classification algorithm; The priority of the comparison algorithm is determined according to the weight coefficient of each feature in the comprehensive feature vector, thereby determining the currently selected comparison algorithm.

6. The vehicle dismantling parts tracing method according to claim 5 is characterized in that: The following steps are also included: Extracting a standard shape feature of each automobile part from the parts database, calculating a first similarity between the shape feature and the standard shape feature according to the comparison algorithm, screening the automobile parts whose first similarity is greater than a preset first similarity threshold from the parts database, and summarizing them into a first candidate parts list; Extracting the standard size feature of each first candidate part from the parts database, calculating the second similarity between the size feature and the standard size feature according to the comparison algorithm, screening out the automobile parts whose second similarity is greater than a preset second similarity threshold from the first candidate parts, and summarizing them into a second candidate parts list; The standard material feature of each second candidate part is extracted from the parts database, the third similarity between the material feature and the standard material feature is calculated according to the comparison algorithm, and the automotive parts whose third similarity is greater than a preset third similarity threshold are screened from the second candidate parts, and they are summarized into a third candidate parts table.

7. The vehicle dismantling parts tracing method according to claim 6 is characterized in that: Determining and outputting the traceability result of the current disassembled parts according to the comparison result specifically includes the following steps: Arrange the third candidate parts list in descending order to obtain a traceability parts list of the current disassembled parts; According to a preset candidate threshold, the automobile parts having a quantity equal to the candidate threshold are selected from top to bottom in the traceability parts list, and the source information corresponding to the selected automobile parts is searched from the parts database; The selected automobile parts and the corresponding source information are summarized into the traceability results of the current disassembled parts and output.

8. A vehicle dismantling parts tracing device, characterized in that: Includes the following modules: An information collection module is used to collect all-round images of the disassembled parts to obtain appearance images, perform material composition analysis on the disassembled parts to obtain material analysis results, and summarize the appearance images and the material analysis results to obtain collection information; An identification code traceability module is used to determine whether the disassembled part has an identification code based on the appearance image, and when the disassembled part has an identification code, search the source information of the current disassembled part from a preset parts database based on the identification code, and when the source information exists, output the traceability result of the current disassembled part based on the source information; A feature extraction module, used for extracting features from the collected information to obtain a comprehensive feature vector when the source information does not exist or the identification code does not exist on the disassembled parts; The comparison and tracing module is used to compare the comprehensive feature vector with the parts database and obtain a comparison result, and determine and output the tracing result of the current disassembled part based on the comparison result.

9. An intelligent terminal, characterized in that: It includes a memory and a processor, wherein the memory stores at least one instruction, at least one program, code set or instruction set, and the at least one instruction, at least one program, code set or instruction set is loaded and executed by the processor to implement the vehicle dismantled parts traceability method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The readable storage medium stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, at least one program, a code set or an instruction set is loaded and executed by a processor to implement the vehicle dismantled parts traceability method as described in any one of claims 1 to 7.

Images