Product disassembly and repair detection method and device, equipment, storage medium and program product

Abstract

The embodiment of the invention provides a product disassembly and repair detection method and device, equipment, a storage medium and a program product. The method comprises the following steps: acquiring an appearance quality inspection report and an X-ray image of an internal structure of a to-be-detected product; performing multi-modal anomaly detection processing on the appearance quality inspection report and the X-ray image to obtain a detection result of the to-be-detected product; wherein the detection result comprises the maintenance processing condition of the to-be-detected product; and if it is determined that the detection result represents that the to-be-detected product has been repaired, performing deep repair processing on the to-be-detected product. The method can improve the efficiency and accuracy of product maintenance detection.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology, and in particular to a product disassembly and repair testing method, apparatus, equipment, storage medium, and program product. Background Technology

[0002] In the process of recycling used electronic products, determining whether a product has been disassembled or repaired is a crucial step affecting recycling pricing and user experience. Signs of disassembly and repair may involve battery replacement, motherboard repair, screen disassembly and reassembly, etc. These actions not only affect the product's performance and lifespan but may also conceal safety hazards.

[0003] In existing technologies, visible components of mobile phones, such as appearance, interfaces, and battery compartments, are inspected manually or using semi-automatic equipment, and surface features such as loose screws, worn casing, and oxidized interfaces are recorded.

[0004] However, the above methods have a high rate of missed detection for concealed repairs (such as motherboard repair and internal component replacement). Summary of the Invention

[0005] This application provides product disassembly and repair testing methods, apparatus, equipment, storage media, and program products, which can improve the efficiency and accuracy of product disassembly and repair testing.

[0006] Firstly, this application provides a product disassembly and repair testing method, including:

[0007] Obtain the appearance quality inspection report and X-ray images of the internal structure of the product to be inspected;

[0008] The appearance quality inspection report and the X-ray image are subjected to multimodal anomaly detection processing to obtain the inspection results of the product to be inspected; wherein, the inspection results include the disassembly and repair process of the product to be inspected;

[0009] If the test results indicate that the product under test has been disassembled and repaired, then the product under test will undergo in-depth disassembly and repair.

[0010] In one possible implementation, the multimodal anomaly detection processing of the appearance inspection report and the X-ray image to obtain the inspection result of the product to be inspected includes:

[0011] The appearance inspection report and the X-ray image are subjected to multimodal anomaly detection processing to obtain the disassembly and repair probability value of the product to be inspected; wherein, the disassembly and repair probability value represents the possibility that the product to be inspected has undergone disassembly and repair processing;

[0012] The test result of the product to be tested is determined based on the disassembly and repair probability value.

[0013] In one possible implementation, the multimodal anomaly detection processing of the appearance inspection report and the X-ray image to obtain the disassembly / repair probability value of the product under inspection includes:

[0014] Attention processing is performed on the appearance quality inspection report to obtain the attention embedding vector of the appearance quality inspection report; wherein, the attention embedding vector represents the appearance features of the product to be inspected;

[0015] Feature extraction is performed on the X-ray image to obtain the image features of the X-ray image; wherein, the image features characterize the internal structural features of the product to be inspected;

[0016] Multimodal anomaly detection processing is performed on the attention embedding vector and the image features to obtain the disassembly and repair probability value of the product to be detected.

[0017] In one possible implementation, the attention processing of the appearance quality inspection report to obtain the attention embedding vector of the appearance quality inspection report includes:

[0018] Based on the category of the product to be tested, information is extracted from the appearance quality inspection report to obtain key information; wherein, the appearance quality inspection report includes the appearance condition and the condition of superficial disassembly and repair.

[0019] Attention processing is performed on the key information to obtain the attention embedding vector of the appearance quality inspection report.

[0020] In one possible implementation, determining the test result of the product to be tested based on the disassembly and repair probability value includes:

[0021] If the disassembly and repair probability value is determined to be greater than or equal to the disassembly and repair threshold, then the test result indicates that the product under test has been disassembled and repaired.

[0022] If the probability value of disassembly and repair is determined to be less than the disassembly and repair threshold, then the test result indicates that the product under test has not undergone disassembly and repair.

[0023] In one possible implementation, the method further includes:

[0024] The disassembly and repair threshold is adjusted based on the product category information of the product to be tested.

[0025] In one possible implementation, the deep disassembly and repair process on the product to be tested includes:

[0026] Based on the abnormal locations and degree of disassembly and repair information in the test results, the product to be tested is subjected to in-depth disassembly and repair.

[0027] Secondly, this application provides a product disassembly and repair testing device, comprising: (characteristic portion)

[0028] The acquisition module is used to acquire the appearance quality inspection report and X-ray images of the internal structure of the product to be inspected.

[0029] The detection module is used to perform multimodal anomaly detection processing on the appearance quality inspection report and the X-ray image to obtain the detection result of the product to be inspected; wherein, the detection result includes the disassembly and repair process of the product to be inspected;

[0030] The processing module is used to perform deep disassembly and repair on the product under test if it is determined that the test result indicates that the product under test has been disassembled and repaired.

[0031] In one possible implementation, the detection module is specifically used to: perform multimodal anomaly detection processing on the appearance quality inspection report and the X-ray image to obtain the disassembly and repair probability value of the product to be inspected; wherein the disassembly and repair probability value characterizes the possibility that the product to be inspected has undergone disassembly and repair processing; and determine the detection result of the product to be inspected based on the disassembly and repair probability value.

[0032] In one possible implementation, the detection module is further specifically configured to: perform attention processing on the appearance quality inspection report to obtain an attention embedding vector of the appearance quality inspection report; wherein the attention embedding vector represents the appearance features of the product to be inspected; perform feature extraction on the X-ray image to obtain image features of the X-ray image; wherein the image features represent the internal structural features of the product to be inspected; and perform multimodal anomaly detection processing on the attention embedding vector and the image features to obtain a disassembly / repair probability value of the product to be inspected.

[0033] In one possible implementation, the detection module is further specifically used to: extract information from the appearance quality inspection report according to the category of the product to be inspected, and obtain key information; wherein, the appearance quality inspection report includes appearance condition and superficial disassembly and repair condition; and perform attention processing on the key information to obtain the attention embedding vector of the appearance quality inspection report.

[0034] In one possible implementation, the detection module is further configured to: if the disassembly probability value is greater than or equal to the disassembly threshold, determine that the detection result indicates that the product under test has undergone disassembly; if the disassembly probability value is less than the disassembly threshold, determine that the detection result indicates that the product under test has not undergone disassembly.

[0035] In one possible implementation, the detection module is further specifically used to: adjust the disassembly and repair threshold according to the category information of the product to be detected.

[0036] In one possible implementation, the processing module is specifically used to: perform in-depth disassembly and repair processing on the product to be tested based on the abnormal location and disassembly and repair information in the test results.

[0037] Thirdly, this application provides an electronic device, including: a memory and a processor;

[0038] The memory stores computer-executed instructions;

[0039] The processor executes computer execution instructions stored in the memory, causing the processor to perform the first aspect and / or various possible implementations of the first aspect as described above.

[0040] Fourthly, this application provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the first aspect and / or various possible embodiments of the first aspect.

[0041] Fifthly, this application provides a computer program product, including a computer program that, when executed by a processor, implements the first aspect and / or various possible implementations of the first aspect.

[0042] The product disassembly and repair testing method, apparatus, equipment, storage medium, and program products provided in this application determine the disassembly and repair status of the product by performing multimodal anomaly detection processing on the appearance quality inspection report and X-ray images of the internal structure of the product to be tested. If it is determined that the product to be tested has been disassembled and repaired, further in-depth disassembly and repair processing is performed on the product to be tested. Furthermore, by introducing multimodal analysis of X-ray images and appearance quality inspection reports, high-precision identification of disassembly and repair traces is achieved, thereby improving the efficiency and accuracy of product disassembly and repair testing. Attached Figure Description

[0043] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0044] Figure 1 This application provides an illustration of an application scenario.

[0045] Figure 2 A schematic flowchart illustrating a product disassembly and repair testing method provided in an embodiment of this application;

[0046] Figure 3 A flowchart illustrating another product disassembly and testing method provided in this application embodiment;

[0047] Figure 4A flowchart illustrating a multimodal mobile phone disassembly and repair testing process provided in this application embodiment;

[0048] Figure 5 This is a schematic diagram of the structure of a product disassembly and repair testing device provided in an embodiment of this application;

[0049] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0050] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0051] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0052] It should be noted that this application can be used in the field of artificial intelligence, or in any field other than artificial intelligence; the application field of this application is not limited.

[0053] Figure 1 This application provides an illustration of an application scenario, such as... Figure 1 As shown, this application is mainly applied to the field of quality inspection of recycled second-hand electronic products, especially the recycling of 3C products such as mobile phones (including computer, communication, and consumer electronic products, such as mobile phones).

[0054] Based on the above scenarios, it is clear that inspecting the visible parts of a product using manual or semi-automatic equipment presents a high rate of missed inspections.

[0055] The product disassembly and repair inspection method provided in this application achieves high-precision identification of disassembly and repair traces by introducing multimodal analysis of X-ray images and appearance quality inspection reports.

[0056] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0057] Figure 2 This is a flowchart illustrating a product disassembly and repair testing method provided in an embodiment of this application, as shown below. Figure 2 As shown, the method includes:

[0058] 201. Obtain the appearance quality inspection report and X-ray images of the internal structure of the product to be tested.

[0059] For example, the execution subject of this embodiment can be an electronic device. The device can obtain an appearance quality inspection report of the product to be inspected, including the appearance inspection results, through user input. The device can also acquire X-ray images of the internal structure of the product to be inspected through an X-ray image acquisition device.

[0060] Among these methods, the appearance inspection results can be obtained by taking pictures of the phone's exterior and then analyzing them, including records of surface features such as loose screws and worn casing.

[0061] For example, for luxury goods, the appearance inspection report includes brand logo wear and watch strap replacement records, and its X-ray image analysis analyzes the integrity of the movement structure; for used cars, the appearance inspection report records paint scratches and interior wear, and its X-ray image analysis analyzes the engine block and chassis structure.

[0062] 202. Perform multimodal anomaly detection processing on the appearance quality inspection report and X-ray images to obtain the test results of the product to be tested; the test results include the disassembly and repair process of the product to be tested.

[0063] For example, the device can call a preset multimodal anomaly detection algorithm, such as a multimodal fusion model, to perform multimodal processing on the text information and X-ray images in the appearance quality inspection report to obtain the inspection results of the product to be inspected; wherein, the inspection results include the disassembly and repair status of the product to be inspected, such as whether there are signs of disassembly and repair or whether it has been disassembled and repaired.

[0064] The multimodal fusion model can be obtained by training an initial model, such as a visual-language model, based on sample data (text information from multiple appearance quality inspection reports and multiple X-ray images from historical time periods). The specific model training method can be set by the user according to the actual situation, and there are no restrictions here.

[0065] In one example, a knowledge graph of repair traces is constructed to establish semantic associations between component status in text reports and structural anomalies in X-ray images. For instance, the text feature of "battery replacement" is matched with anomaly regions of battery compartment welding traces in X-ray images. By enhancing the semantic associations of cross-modal features through the knowledge graph, the false negative rate of covert repairs is further reduced.

[0066] 203. If the test results indicate that the product under test has been disassembled and repaired, then the product under test shall be subjected to in-depth disassembly and repair.

[0067] For example, based on the test results of the product to be tested, if it is determined that the product to be tested has undergone disassembly and repair, further in-depth disassembly and repair can be carried out on the product to further inspect the internal structure of the product to be tested. If it is determined that the product to be tested has not undergone disassembly and repair, the product to be tested is then put into storage.

[0068] This embodiment provides a product disassembly and repair inspection method. By introducing multimodal analysis of X-ray images and appearance quality inspection reports, it enables non-destructive testing of product disassembly and repair issues and high-precision identification of disassembly and repair traces, thereby improving the efficiency and accuracy of product disassembly and repair inspection.

[0069] Figure 3 A flowchart illustrating another product disassembly and testing method provided in this application embodiment is shown below. Figure 3 As shown, the method includes:

[0070] 301. Obtain the appearance quality inspection report and X-ray images of the internal structure of the product to be tested.

[0071] For example, this step can be referred to as step 201, which will not be repeated here.

[0072] 302. Perform multimodal anomaly detection processing on the appearance quality inspection report and X-ray image to obtain the disassembly and repair probability value of the product to be inspected; wherein, the disassembly and repair probability value represents the possibility that the product to be inspected has undergone disassembly and repair processing.

[0073] For example, the device can invoke a preset multimodal anomaly detection algorithm, such as a hybrid architecture of Transformer and Convolutional Neural Network (Transformer-CNN), to perform multimodal processing on the text information and X-ray images in the appearance quality inspection report. Specifically, this may include multimodal feature alignment, feature processing, normalization, and other operations to predict the disassembly and repair probability value of the product to be inspected. The disassembly and repair probability value represents the likelihood that the product to be inspected has undergone disassembly and repair.

[0074] In one example, a multimodal anomaly detection algorithm can be a dynamic detection model based on time-series data. Specifically, historical disassembly and repair records (such as "screen replaced in 2023") can be added to the appearance quality inspection report. The temporal patterns of disassembly and repair activities can be analyzed using a time-series model (such as a Long Short-Term Memory Network, LSTM). For example, frequent replacement of the same component may indicate a structural problem with the equipment. By identifying potential repeated disassembly and repair activities through time-series analysis, the ability to detect maliciously refurbished equipment can be improved.

[0075] In one possible implementation, step 302 includes the following steps:

[0076] The first step is to perform attention processing on the appearance quality inspection report to obtain the attention embedding vector of the appearance quality inspection report; whereby the attention embedding vector represents the appearance features of the product to be inspected.

[0077] The second step is to extract features from the X-ray image to obtain the image features of the X-ray image; among them, the image features represent the internal structural features of the product to be inspected.

[0078] The third step involves performing multimodal anomaly detection processing on the attention embedding vector and image features to obtain the disassembly and repair probability value of the product to be detected.

[0079] Specifically, the device invokes a preset multimodal anomaly detection algorithm, such as a Transformer-CNN hybrid architecture. Based on the Transformer layer of this architecture, attention processing is applied to the appearance inspection report of the product to be inspected, resulting in an attention embedding vector representing the appearance features of the product, such as "loose screws" or "battery replacement". Based on the CNN layer of this architecture, feature extraction is performed on the X-ray image of the product to be inspected, resulting in image features representing the internal structural features of the product, such as motherboard solder marks and component displacement. Based on the feature fusion layer of this architecture, multimodal feature fusion processing is performed on the obtained attention embedding vector and image features, such as attention mechanisms (e.g., self-attention or multi-head attention) to align cross-modal feature processing, resulting in a repair probability value for the product to be inspected.

[0080] In one example, for multi-scale feature extraction of X-ray images, multi-scale convolutional kernels (such as 3×3 and 5×5) can be introduced into the CNN to extract local details (such as the edge of screw holes) and global structure (such as motherboard layout) features at the same time. Multi-scale analysis can improve the detection capability of tiny disassembly and repair traces (such as 0.1mm-level solder joint anomalies).

[0081] By fusing multimodal data, the need for manual disassembly is reduced, thus lowering labor costs and the risk of equipment damage, and further improving detection efficiency. By aligning cross-modal features, the rate of missed detections in concealed disassembly and repair is reduced, thereby enhancing detection accuracy.

[0082] In one possible implementation, the first step includes:

[0083] Step 1: Extract information from the appearance inspection report based on the product category to obtain key information; the appearance inspection report includes the appearance condition and superficial disassembly and repair details.

[0084] Step 2: Perform attention processing on key information to obtain the attention embedding vector of the appearance quality inspection report.

[0085] Specifically, the product to be inspected is identified to determine its category, such as used cars, mobile phones, computers, and luxury goods. Based on the product category, information is extracted from the text information in the product's appearance quality inspection report to obtain key information to match the inspection requirements of different categories. Using the attention mechanism of a multimodal anomaly detection algorithm, this key information is processed to obtain the corresponding attention embedding vector.

[0086] The appearance inspection report includes both the physical condition and the details of any minor repairs. The details of minor repairs can be obtained through manual or semi-automatic inspection of components such as the battery compartment and interfaces, and include a text report showing the condition of visible components and replacement records.

[0087] By extracting information and processing attention from appearance inspection reports based on the product category to be inspected, this solution can be quickly migrated to inspection scenarios for high-value items such as luxury goods and used cars, improving the adaptability of cross-category applications.

[0088] 303. Determine the test results of the product to be tested based on the disassembly and repair probability value.

[0089] For example, by using preset judgment logic, the probability value of disassembly and repair of the product to be tested is analyzed to obtain the test result of the product to be tested, so as to determine whether the product to be tested has traces of disassembly and repair or has been disassembled and repaired.

[0090] In one example, the confidence level of the disassembly and repair probability value is calculated according to a preset calibration function. If the confidence level is determined to be greater than a preset threshold, the disassembly and repair probability value is determined to be reliable. Then, according to a decoding algorithm, such as a conditional random field algorithm, the disassembly and repair probability value is decoded to obtain the corresponding disassembly and repair label, including whether the product under test has been disassembled and repaired or whether the product under test has not been disassembled and repaired.

[0091] By designing a dual-track grading strategy, the decision on whether manual deep dismantling is needed is made in a tiered manner based on the dismantling and repair probability value output by the multimodal model, thereby further reducing the dependence on deep dismantling and reducing labor costs and equipment damage risks.

[0092] In one possible implementation, step 303 includes: if the probability value of disassembly / repair is determined to be greater than or equal to a disassembly / repair threshold, then the test result indicates that the product under test has undergone disassembly / repair processing. If the probability value of disassembly / repair is determined to be less than a disassembly / repair threshold, then the test result indicates that the product under test has not undergone disassembly / repair processing.

[0093] Specifically, a preset disassembly / repair threshold is invoked, and this threshold is compared with the disassembly / repair probability value of the product to be tested. If the disassembly / repair probability value is greater than or equal to the threshold, the product to be tested is determined to have undergone disassembly / repair; if the disassembly / repair probability value is less than the threshold, the product to be tested is determined not to have undergone disassembly / repair.

[0094] By using multimodal data fusion and a dual-track classification strategy, we can further reduce manual disassembly operations, improve disassembly and repair testing efficiency, reduce the missed detection rate, and improve testing accuracy.

[0095] In one possible implementation, step 303 further includes: adjusting the disassembly and repair threshold according to the category information of the product to be tested.

[0096] Specifically, the product to be tested is identified to determine its category, such as used cars, mobile phones, computers, and luxury goods. Based on the category of the product to be tested, the preset disassembly and repair threshold is adjusted to accommodate the testing needs of different categories (e.g., the difference in sensitivity to disassembly and repair marks between mobile phones and luxury goods).

[0097] In one example, a dynamic threshold adjustment mechanism is introduced to dynamically adjust the disassembly and repair threshold based on the product category and historical data distribution. Through dynamic threshold optimization, the adaptability of the detection model to different product categories is improved, the false positive and false negative rates are reduced, and the detection accuracy is further enhanced.

[0098] 304. If the test results indicate that the product under test has been disassembled and repaired, then based on the abnormal location and degree of disassembly and repair information in the test results, the product under test shall be subjected to in-depth disassembly and repair.

[0099] For example, based on the test results of the product under test, if it is determined that the product under test has been disassembled and repaired, information is extracted from the test results to obtain information on the abnormal locations and extent of disassembly and repair marks on the product under test. The extent of disassembly and repair information represents the degree of disassembly and repair on the product under test. Based on the abnormal locations and extent of disassembly and repair marks on the product under test, in-depth disassembly and repair processing is performed on the product under test; for example, the abnormal locations and extent of disassembly and repair marks on the product under test are visualized to instruct the user to perform manual in-depth disassembly and repair processing.

[0100] For example, Figure 4 This application provides a flowchart illustrating a multimodal mobile phone disassembly and repair testing process, as shown in the embodiments below. Figure 4 As shown, the phone undergoes appearance inspection and shallow disassembly, and the two inspection results are used to generate corresponding appearance quality inspection reports. We use a trained large model algorithm to extract key information from the reports and generate attention embedding vectors of specific dimensions. At the same time, we use an Xray machine to take pictures of the phone to obtain Xray images. Then, the attention embedding vectors and Xray images are input together into a trained multimodal anomaly detection algorithm for algorithm inference. The algorithm determines whether the phone has been disassembled and repaired. For phones that have been disassembled and repaired, the abnormal location and score (disassembly and repair degree information) are output, and the results are confirmed by manual deep disassembly. Phones that have not been disassembled and repaired can be directly added to the database.

[0101] By introducing multimodal analysis of X-ray images and appearance reports, the reliance on manual deep disassembly can be reduced, and users can be instructed to perform manual deep disassembly and repair based on the inspection results, which can further reduce labor costs and equipment damage risks.

[0102] In this embodiment, based on the above embodiments, on the one hand, high-precision identification of disassembly and repair traces is achieved by combining text (appearance report) and image (X-ray) data for collaborative analysis; on the other hand, through joint reasoning of appearance quality inspection report and X-ray image, layered decision-making is made on whether manual deep disassembly is required, reducing reliance on manual deep disassembly and improving detection efficiency and adaptability.

[0103] Figure 5 This is a schematic diagram of the structure of a product disassembly and repair testing device provided in an embodiment of this application, as shown below. Figure 5 As shown, the device includes:

[0104] The acquisition module 401 is used to acquire the appearance quality inspection report and X-ray images of the internal structure of the product to be inspected.

[0105] The detection module 402 is used to perform multimodal anomaly detection processing on the appearance quality inspection report and X-ray image to obtain the detection results of the product to be inspected; the detection results include the disassembly and repair process of the product to be inspected.

[0106] The processing module 403 is used to perform deep disassembly and repair on the product to be tested if the test result indicates that the product to be tested has been disassembled and repaired.

[0107] In one possible implementation, the detection module 402 is specifically used to: perform multimodal anomaly detection processing on the appearance quality inspection report and X-ray image to obtain the disassembly and repair probability value of the product to be inspected; wherein, the disassembly and repair probability value characterizes the possibility that the product to be inspected has undergone disassembly and repair processing; and determine the inspection result of the product to be inspected based on the disassembly and repair probability value.

[0108] In one possible implementation, the detection module 402 is further specifically used for: performing attention processing on the appearance quality inspection report to obtain the attention embedding vector of the appearance quality inspection report; wherein the attention embedding vector represents the appearance features of the product to be inspected; performing feature extraction on the X-ray image to obtain the image features of the X-ray image; wherein the image features represent the internal structural features of the product to be inspected; and performing multimodal anomaly detection processing on the attention embedding vector and the image features to obtain the disassembly and repair probability value of the product to be inspected.

[0109] In one possible implementation, the detection module 402 is further specifically used to: extract information from the appearance quality inspection report according to the category of the product to be inspected, and obtain key information; wherein, the appearance quality inspection report includes the appearance condition and the shallow disassembly and repair condition; and perform attention processing on the key information to obtain the attention embedding vector of the appearance quality inspection report.

[0110] In one possible implementation, the detection module 402 is further specifically used to: if the probability value of disassembly and repair is determined to be greater than or equal to the disassembly and repair threshold, then determine that the detection result indicates that the product under test has been disassembled and repaired; if the probability value of disassembly and repair is determined to be less than the disassembly and repair threshold, then determine that the detection result indicates that the product under test has not been disassembled and repaired.

[0111] In one possible implementation, the detection module 402 is also specifically used to: adjust the disassembly and repair threshold according to the category information of the product to be detected.

[0112] In one possible implementation, the processing module 403 is specifically used to: perform in-depth disassembly and repair processing on the product to be tested based on the abnormal location and disassembly and repair information in the test results.

[0113] The apparatus in this embodiment can execute the technical solutions in the above method. Its specific implementation process and technical principles are the same, and will not be repeated here.

[0114] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application, such as... Figure 6As shown, the electronic device includes: a memory 501 and a processor 502; the memory 501 is a memory used to store instructions executable by the processor 502.

[0115] The processor 502 is configured to perform the method provided in the above embodiments.

[0116] The electronic device also includes a receiver 503 and a transmitter 504. The receiver 503 is used to receive instructions and data sent by other devices, and the transmitter 504 is used to send instructions and data to external devices.

[0117] The specific implementation process of the processor can be found in the above method embodiments, and its implementation principle and technical effect are similar, so it will not be repeated here.

[0118] In the above embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0119] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed on a computer, cause the computer to perform the technical solutions described above.

[0120] The aforementioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory, electrically erasable programmable read-only memory, erasable programmable read-only memory, programmable read-only memory, read-only memory, magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.

[0121] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. The readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an application-specific integrated circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components in a device.

[0122] This application also provides a computer program product, which includes a computer program stored in a computer-readable storage medium. At least one processor can read the computer program from the computer-readable storage medium, and when the at least one processor executes the computer program, it can implement the technical solutions in the above embodiments.

[0123] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0124] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as magnetic disks or optical disks.

[0125] Finally, it should be noted that other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein, and is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of the invention is limited only by the appended claims.

Claims

1. A product disassembly and repair testing method, characterized in that, include: Obtain the appearance quality inspection report and X-ray images of the internal structure of the product to be inspected; The appearance quality inspection report and the X-ray image are subjected to multimodal anomaly detection processing to obtain the inspection results of the product to be inspected; wherein, the inspection results include the disassembly and repair process of the product to be inspected; If the test results indicate that the product under test has been disassembled and repaired, then the product under test will undergo in-depth disassembly and repair.

2. The method according to claim 1, characterized in that, The process of performing multimodal anomaly detection processing on the appearance quality inspection report and the X-ray image to obtain the inspection results of the product to be inspected includes: The appearance inspection report and the X-ray image are subjected to multimodal anomaly detection processing to obtain the disassembly and repair probability value of the product to be inspected; wherein, the disassembly and repair probability value represents the possibility that the product to be inspected has undergone disassembly and repair processing; The test result of the product to be tested is determined based on the disassembly and repair probability value.

3. The method according to claim 2, characterized in that, The process of performing multimodal anomaly detection processing on the appearance inspection report and the X-ray image to obtain the disassembly and repair probability value of the product under inspection includes: Attention processing is performed on the appearance quality inspection report to obtain the attention embedding vector of the appearance quality inspection report; wherein, the attention embedding vector represents the appearance features of the product to be inspected; Feature extraction is performed on the X-ray image to obtain the image features of the X-ray image; wherein, the image features characterize the internal structural features of the product to be inspected; Multimodal anomaly detection processing is performed on the attention embedding vector and the image features to obtain the disassembly and repair probability value of the product to be detected.

4. The method according to claim 3, characterized in that, The attention processing of the appearance quality inspection report to obtain the attention embedding vector of the appearance quality inspection report includes: Based on the category of the product to be tested, information is extracted from the appearance quality inspection report to obtain key information; wherein, the appearance quality inspection report includes the appearance condition and the condition of superficial disassembly and repair. Attention processing is performed on the key information to obtain the attention embedding vector of the appearance quality inspection report.

5. The method according to claim 2, characterized in that, The step of determining the test result of the product to be tested based on the disassembly and repair probability value includes: If the disassembly and repair probability value is determined to be greater than or equal to the disassembly and repair threshold, then the test result indicates that the product under test has been disassembled and repaired. If the probability value of disassembly and repair is determined to be less than the disassembly and repair threshold, then the test result indicates that the product under test has not undergone disassembly and repair.

6. The method according to claim 5, characterized in that, The method further includes: The disassembly and repair threshold is adjusted based on the product category information of the product to be tested.

7. The method according to any one of claims 1-6, characterized in that, The deep disassembly and repair process of the product to be tested includes: Based on the abnormal locations and degree of disassembly and repair information in the test results, the product to be tested is subjected to in-depth disassembly and repair.

8. A product disassembly and repair testing device, characterized in that, include: The acquisition module is used to acquire the appearance quality inspection report and X-ray images of the internal structure of the product to be inspected. The detection module is used to perform multimodal anomaly detection processing on the appearance quality inspection report and the X-ray image to obtain the detection result of the product to be inspected; wherein, the detection result includes the disassembly and repair process of the product to be inspected; The processing module is used to perform deep disassembly and repair on the product under test if it is determined that the test result indicates that the product under test has been disassembled and repaired.

9. An electronic device, characterized in that, include: Memory, processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory, causing the processor to perform the method as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1-7.

11. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the method described in any one of claims 1-7.

Images