Communication terminal full-dimension detection method and system based on end-edge-cloud cooperation

Abstract

The application relates to the technical field of communication terminal automatic detection, and provides a communication terminal full-dimension detection method and system based on end-edge cloud cooperation, which comprises the following steps: a previous mechanical arm is controlled to move a current communication terminal to a previous detection station, image recognition is performed on terminal surface images returned by visual sensors of the previous detection station, and an image recognition result is obtained; a standard detection strategy benchmark corresponding to the model of the current communication terminal is planned based on the image recognition result, a subsequent mechanical arm of a subsequent detection station is executed to perform full-dimension function test, and function test data are obtained; state mapping is performed based on the image recognition result, the function test data and a three-dimensional digital model of the current communication terminal, and full-dimension state mapping results are obtained; and the full-dimension state mapping results are matched with a recycling price database, and recycling evaluation results of the current communication terminal are obtained. The application improves the detection accuracy and the price evaluation accuracy in the communication terminal detection process.

Description

Technical Field

[0001] This invention relates to the field of automated testing technology for communication terminals, and in particular to a method and system for full-dimensional testing of communication terminals based on end-edge-cloud collaboration. Background Technology

[0002] With the widespread adoption of mobile communication technology, the market for the recycling and circulation of communication terminals (such as smartphones) is expanding rapidly. In the secondhand trading process, objective and accurate quality testing and value assessment of communication terminals are essential to ensuring fairness in transactions.

[0003] Existing automated testing systems for communication terminals typically employ a multi-station operation mode. First, a robotic arm moves the terminal to be tested to a visual inspection station, where visual sensors capture images of the terminal's surface, identifying and recording the presence of physical defects such as scratches and cracks. Next, the terminal is moved to a functional testing station, where a full functional traversal test is performed on the terminal's screen, camera, audio, and other modules according to a pre-set standard testing procedure, and the test data for each module is recorded. Finally, the visual inspection results and functional test results are combined, and the terminal's recycling valuation price is determined through table lookup or weighted calculation.

[0004] However, in practical applications, it has been found that the recycling assessment results obtained by using the existing detection methods often deviate from the actual condition of the terminal. For example, for some terminals with special appearance defects but passing functional tests, or terminals with intact appearance but minor differences in internal performance, the assessment results given by the existing methods are prone to overestimation or underestimation, making it difficult to accurately reflect the actual remaining value of the terminal, thereby reducing the detection accuracy of the communication terminal detection process and the accuracy of price assessment. Summary of the Invention

[0005] This invention provides a method and system for full-dimensional detection of communication terminals based on edge-cloud collaboration, aiming to improve the detection accuracy and price evaluation accuracy of communication terminals.

[0006] In a first aspect, the present invention provides a method for full-dimensional detection of communication terminals based on edge-cloud collaboration, applied to a full-dimensional detection system for communication terminals integrating a robotic arm, edge nodes, and a server cluster; the robotic arm is used to perform physical transfer of the communication terminal, the edge nodes are used to perform production line scheduling, and the server cluster is used to provide a three-dimensional digital model, a standard detection strategy benchmark, and a recycling price database; the method includes: The preceding robotic arm is controlled to move the current communication terminal to the preceding inspection station. The visual sensor of the preceding inspection station returns the terminal surface image, and image recognition is performed to obtain the image recognition result. The image recognition result indicates whether there is a defect area on the terminal surface. Based on the image recognition results and the standard detection strategy benchmark corresponding to the current communication terminal model, the subsequent robotic arm of the subsequent detection station performs full-dimensional functional testing to obtain functional test data. Based on the image recognition results and the functional test data, combined with the current three-dimensional digital model of the communication terminal, a state mapping is performed to obtain a full-dimensional state mapping result. The recycling assessment result of the current communication terminal is obtained by matching the full-dimensional state mapping result with the recycling price database.

[0007] Secondly, the present invention also provides a communication terminal full-dimensional detection system based on end-edge-cloud collaboration, used to implement the communication terminal full-dimensional detection method based on end-edge-cloud collaboration as described in the first aspect; the system includes: The image recognition module is used to control the preceding robotic arm to move the current communication terminal to the preceding inspection station, perform image recognition on the terminal surface image returned by the vision sensor of the preceding inspection station, and obtain the image recognition result; the image recognition result indicates whether there is a defect area on the terminal surface. The full-dimensional functional testing module is used to plan the subsequent robotic arm of the subsequent testing station based on the image recognition results and the standard detection strategy benchmark corresponding to the current communication terminal model to perform full-dimensional functional testing and obtain functional test data. The full-dimensional state mapping module is used to perform state mapping based on the image recognition results and the functional test data, combined with the current three-dimensional digital model of the communication terminal, to obtain the full-dimensional state mapping result. The recycling assessment module is used to match the full-dimensional state mapping results with the recycling price database to obtain the recycling assessment result of the current communication terminal.

[0008] Thirdly, the present invention also provides an electronic device, comprising: a memory for storing a computer program; and a processor for reading and executing the computer program, thereby realizing the communication terminal full-dimensional detection method based on edge-cloud collaboration as described above.

[0009] Fourthly, the present invention also provides a non-transitory computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-described end-edge-cloud collaborative communication terminal full-dimensional detection method.

[0010] Fifthly, the present invention provides a computer program product, including a computer program, which, when executed by a processor, implements the above-mentioned method for full-dimensional detection of communication terminals based on edge-cloud collaboration.

[0011] The edge-cloud collaborative communication terminal detection method provided in this invention accurately captures the true condition of defective areas on the terminal surface through terminal transfer and image recognition. It overcomes the limitations of simple appearance inspection that merely records the presence or absence of defects. By combining standard detection strategies corresponding to the specific model, it plans subsequent full-dimensional functional testing processes, replacing the fixed full-scale traversal testing mode. It can adapt refined testing logic for terminals with special appearance defects and terminals with intact appearances, effectively capturing subtle internal performance differences and generating accurate functional test data. This solves the problems of being unable to correlate appearance status with functional testing, easily overlooking internal performance issues, or misjudging defects. To address the shortcomings of traditional methods, this approach relies on a 3D digital model of the terminal to perform a full-dimensional state mapping between precise image recognition appearance data and refined functional test data. This enables a three-dimensional and comprehensive restoration of the terminal's physical appearance and internal performance, overcoming the limitation of simply summarizing two independent test data sets and failing to accurately match the terminal's true remaining value. Finally, the value assessment is completed by accurately matching the full-dimensional state mapping results with the recycling price database. This improves upon the problem of overestimation or underestimation of terminals with special defects or subtle performance differences in existing technologies, effectively enhancing the accuracy of both the testing process and price assessment in the communication terminal testing process. Attached Figure Description

[0012] Figure 1 This is a flowchart illustrating the multi-dimensional detection method for communication terminals based on end-edge-cloud collaboration provided in an embodiment of the present invention. Figure 2 This is a schematic diagram of the structure of a communication terminal full-dimensional detection system based on edge-cloud collaboration provided in an embodiment of the present invention; Figure 3 An embodiment diagram of the electronic device provided in this invention; Figure 4 An embodiment diagram of a computer-readable storage medium provided in accordance with the present invention. Detailed Implementation

[0013] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0014] Optionally, Figure 1This is a flowchart illustrating the end-edge-cloud collaborative full-dimensional detection method for communication terminals provided by this invention. The method is applied to a full-dimensional detection system for communication terminals, integrating a robotic arm, edge nodes, and a server cluster. The robotic arm performs the physical transfer and detection of the communication terminals, the edge nodes perform production line scheduling, and the server cluster performs data aggregation and value assessment. The end-edge-cloud collaborative full-dimensional detection method for communication terminals includes: Step 10: Control the preceding robotic arm to move the current communication terminal to the preceding detection station, perform image recognition on the terminal surface image returned by the vision sensor of the preceding detection station, and obtain the image recognition result.

[0015] Optionally, the communication terminal full-dimensional detection system in this embodiment of the invention has multiple independent detection stations of different dimensions, including but not limited to detection station A (appearance and vision), detection station B (radio frequency communication), detection station C (acoustic audio), detection station D (interface electrical) and detection station E (screen display). Each detection station is equipped with a corresponding six-axis industrial robotic arm as a handling mechanism (with a micro touch probe integrated at the end).

[0016] The edge node sends a transfer control command to the preceding robotic arm that performs the loading action. The preceding robotic arm, according to the transfer control command, grasps the current communication terminal from the first buffer and places it on the positioning fixture at the preceding inspection station. The preceding inspection station is a visual inspection station, equipped with a vision sensor, including a high-resolution industrial camera and a ring-shaped supplementary light source. The high-resolution industrial camera captures images of the six surfaces of the current communication terminal at a preset shooting angle and distance, obtaining six terminal surface images. These six terminal surface images correspond to the front, back, left side, right side, top, and bottom surfaces of the current communication terminal, respectively.

[0017] The edge node receives six terminal surface images returned by the vision sensor and inputs each of the six terminal surface images into a pre-trained defect detection model for image recognition. The defect detection model is an image segmentation model built on a convolutional neural network. The output of the defect detection model is a binary mask image with the same size as the input terminal surface images. In the binary mask image, the regions with a pixel value of 1 correspond to defect regions, and the regions with a pixel value of 0 correspond to non-defect regions.

[0018] Edge nodes perform connected component analysis on the binary mask image corresponding to each terminal surface image. If a connected component with a pixel value of 1 exists in the binary mask image, it is determined that a defective region exists on the surface of the current communication terminal corresponding to that terminal surface image. The edge nodes then extract the defect contour coordinates, defect area, and defect location identifier of the defective region from the binary mask image. The defect location identifier indicates the surface to which the defective region belongs among the six surfaces of the current communication terminal. The edge nodes summarize the determination results, defect contour coordinates, defect area, and defect location identifiers for the six terminal surface images into an image recognition result. If no connected component with a pixel value of 1 exists in any of the six binary mask images corresponding to the terminal surface images, the image recognition result indicates that no defective region exists on the surface of the current communication terminal, and the image recognition result includes a defect-free determination identifier.

[0019] Step 20: Based on the image recognition results and the standard detection strategy benchmark corresponding to the current communication terminal model, the subsequent robotic arm of the subsequent detection station performs full-dimensional functional testing to obtain functional test data.

[0020] Optionally, the server cluster retrieves the standard testing strategy benchmark corresponding to the current communication terminal model from the standard testing strategy benchmark library. The standard testing strategy benchmark defines the standard parameter range and standard test procedure that each functional module of the model should meet under normal conditions.

[0021] The edge node determines whether the defective region covers the area of ​​a specific functional module of the current communication terminal based on the surface location of the defective region in the image recognition results. If the defective region covers the area of ​​the specific functional module, an enhanced test strategy is generated for that specific functional module based on the standard detection strategy benchmark. The enhanced test strategy increases the number of test items or improves the test accuracy level compared to the standard test procedure. If the defective region does not cover the area of ​​the specific functional module, the test of the corresponding functional module is directly executed according to the standard test procedure defined in the standard detection strategy benchmark. The edge node arranges the enhanced test strategy and the standard test procedure into a test task sequence for subsequent test stations, and assigns the test task sequence to the corresponding subsequent test stations according to the equipment type of each subsequent test station. The subsequent robotic arm of the subsequent test station performs full-dimensional functional testing on the current communication terminal according to the assigned test task sequence. Each subsequent test station reports the raw data collected during the test, and the edge node summarizes the raw data reported by each subsequent test station into functional test data, as detailed in steps 201 to 206.

[0022] The standard testing strategy benchmark refers to a set of standardized test schemes pre-stored in the server cluster, categorized by communication model. Each model's standard testing strategy benchmark includes a list of standard test items for each functional module of that communication terminal model, the standard test procedure for each test item, and the standard parameter range that each test item should meet under normal conditions. Full-dimensional functional testing refers to the process of testing the current communication terminal's RF communication function, acoustic audio function, interface electrical function, and screen display function separately. Full-dimensional functional testing covers testing stations B, C, D, and E.

[0023] Enhanced testing strategies refer to test plans that, based on standard testing procedures, add extra test items or improve test sampling accuracy for specific functional modules that may be impaired due to defect area coverage. Enhanced testing strategies involve more test items than standard testing procedures. Functional test data refers to the data set that characterizes the actual operating status of each functional module of the current communication terminal, collected and reported by each subsequent testing station during the execution of full-dimensional functional testing.

[0024] Step 30: Based on the image recognition results and functional test data, combined with the current three-dimensional digital model of the communication terminal, state mapping is performed to obtain the full-dimensional state mapping result.

[0025] Optionally, the server cluster retrieves a 3D digital model corresponding to the current communication terminal from the 3D digital model library based on the current communication terminal model. The 3D digital model is a digital representation of the geometric structure and spatial layout of the internal functional modules of the current communication terminal.

[0026] Edge nodes map the defect contour coordinates and defect location markers from the image recognition results to the corresponding surface locations in the 3D digital model, marking the spatial distribution of the defect area in the 3D digital model. They also associate various performance indicators from the functional test data with the spatial locations of corresponding functional modules in the 3D digital model, establishing a correspondence between the functional test data and each functional module in the 3D digital model. Based on the spatial distribution of the defect area in the 3D digital model, the edge nodes calculate the spatial distance between the defect area and each functional module. If the spatial distance between a functional module and the defect area is less than a preset distance threshold, the functional test data corresponding to that module is marked as defect-affected data. If the spatial distance between a functional module and the defect area is greater than or equal to the distance threshold, the functional test data corresponding to that module is marked as independent data. Based on the marked spatial distribution of the defect area, the functional test data associated with each functional module, the defect-affected data, and the independent data in the 3D digital model, the edge nodes determine the full-dimensional state mapping result, as detailed in steps 301 to 306.

[0027] Among them, the three-dimensional digital model refers to the three-dimensional geometric model that is pre-stored in the server cluster and corresponds one-to-one with the communication model. The three-dimensional digital model includes the outer shell geometric dimensions of the communication terminal of that model, the coordinate definition of each surface, and the three-dimensional spatial coordinates and occupied volume of each internal functional module.

[0028] Optionally, the output of the preceding detection station is provided with a first buffer. The first buffer is used to temporarily store communication terminals that have completed the preceding detection station's test. A first area status sensor is provided in the first buffer to detect whether a communication terminal is stored in the first buffer. The first area status includes an occupied state and an idle state. An occupied state indicates that there is a communication terminal in the first buffer that has completed the preceding test but has not yet been transferred to the subsequent detection station. An idle state indicates that there is no communication terminal in the first buffer. The input of the subsequent detection station is provided with a second buffer. The second buffer is used to temporarily store communication terminals waiting to enter the subsequent detection station. A second area status sensor is provided in the second buffer to detect whether a communication terminal is stored in the second buffer. The second area status includes an occupied state and an idle state. An occupied state indicates that there is a communication terminal in the second buffer waiting to enter the subsequent detection station. An idle state indicates that there is no communication terminal in the second buffer.

[0029] The edge node reads the status signals of the first and second area status sensors in real time at a preset monitoring cycle. When it detects that the current communication terminal has completed testing at the preceding detection station, it reads the status of the second area in the second buffer. If the second area is idle, a release command is sent to the preceding robotic arm used to transfer the current communication terminal. The release command includes the target position coordinates, which are the placement coordinates of the second buffer. The preceding robotic arm grabs the current communication terminal from the positioning fixture at the current detection station according to the release command and transfers the current communication terminal to the placement coordinates of the second buffer. If the second area is occupied, the release command is not sent temporarily. The preceding robotic arm keeps the current communication terminal on the positioning fixture at the preceding detection station and waits until the edge node detects that the second area has become idle in the next monitoring cycle, and then sends the release command.

[0030] When a pending communication terminal is detected in the first buffer, the edge node reads the status of the first area of ​​the preceding detection station. If the first area is in an idle state, it means that the preceding detection station is not currently performing a test task. The edge node then sends a grasping command to the loading robot arm. This command includes source and target position coordinates. The source coordinates are the coordinates where the pending communication terminal is stored in the first buffer, and the target coordinates are the coordinates of the positioning fixture at the preceding detection station. The loading robot arm grasps the next communication terminal from the first buffer according to the grasping command and moves it to the positioning fixture at the preceding detection station. If the first area is in an occupied state, it means that the preceding detection station is currently performing a test task. The edge node does not send a grasping command and waits until the edge node detects that the first area has become idle in the next monitoring cycle before sending the grasping command.

[0031] In one embodiment, detection station A is the preceding detection station, and detection station B is the following detection station. Assuming that a smartphone of model M has completed visual inspection at detection station A, the edge node detects that the test at detection station A is complete. The edge node reads the second region state of the second buffer corresponding to the input end of detection station B. Assuming the second region state is idle, it sends a release command to the preceding robotic arm corresponding to detection station A. The target position coordinates in the release command are the placement coordinates of the second buffer. The preceding robotic arm picks up the smartphone of model M from the positioning fixture at detection station A and transfers it to the second buffer. Simultaneously, the edge node detects that there is one smartphone N to be processed in the first buffer corresponding to the output end of detection station A. The edge node reads the first region state of detection station A. Assuming the first region state is idle, it sends a gripping command to the robotic arm used for loading. The source position coordinates in the gripping command are the storage coordinates of smartphone N in the first buffer, and the target position coordinates are the coordinates of the positioning fixture at detection station A. The loading robotic arm picks up smartphone N from the first buffer and transfers it to the positioning fixture at detection station A.

[0032] Step 40: Match the full-dimensional state mapping results with the recycling price database to obtain the recycling assessment results of the current communication terminal.

[0033] Optionally, the server cluster stores a recycling price database. This database is categorized by communication model and stores the benchmark recycling price for each model of communication terminal under different status levels, along with the status determination rules corresponding to each status level. The status determination rules include deduction rules for appearance defects and deduction rules for functional deviations. The appearance defect deduction rules define the price deduction ratios corresponding to different defect area ranges and different defect locations. The functional deviation deduction rules define the price deduction ratios corresponding to different degrees when functional test data deviates from the standard parameter range.

[0034] Edge nodes upload the full-dimensional state mapping results to the server cluster. The server cluster extracts the defect area, defect location identifier, functional test data of each functional module, and defect-affected or independent markers from the full-dimensional state mapping results. Based on the current communication terminal model, the server cluster retrieves the benchmark recycling price corresponding to that model from the recycling price database.

[0035] According to the appearance defect deduction rules, the server cluster calculates the price deduction ratio corresponding to the appearance defect based on the defect area range and defect location identifier of the defect area. The base recycling price is multiplied by the price deduction ratio corresponding to the appearance defect to obtain the appearance defect deduction amount.

[0036] According to the functional deviation deduction rules, the server cluster calculates the degree of deviation between the functional test data and the corresponding standard parameter range for each functional test data. For functional test data marked as affected by defects, the final deviation degree is obtained by multiplying the deviation degree by a preset influence amplification factor. For functional test data marked as independent data, the deviation degree is directly used as the final deviation degree.

[0037] The server cluster determines the price deduction ratio corresponding to each functional test data point from the functional deviation deduction rules based on the deviation range into which the final deviation of each functional test data point falls. The total functional deviation deduction ratio is obtained by summing the price deduction ratios corresponding to each functional test data point. The functional deviation deduction amount is then obtained by multiplying the benchmark recovery price by the total functional deviation deduction ratio.

[0038] The server cluster subtracts the deductions for appearance defects and functional deviations from the benchmark recycling price to obtain the current recycling assessment price of the communication terminal. The recycling assessment price, along with the corresponding deductions for appearance defects, functional deviations, and other deduction details, are packaged into the recycling assessment result.

[0039] The embodiments of the present invention improve the detection accuracy of the communication terminal detection process and the accuracy of price evaluation.

[0040] Optionally, steps 201 to 206 include: Step 201: If the image recognition result indicates that there is no defect area, then a full detection execution instruction is generated based on the standard physical detection trajectory and standard test parameter configuration in the standard detection strategy benchmark.

[0041] Optionally, the edge node reads the judgment identifier from the image recognition result. If the judgment identifier indicates no defect, it sends a policy retrieval request to the server cluster, carrying the terminal model of the current communication terminal. The server cluster retrieves the standard detection policy benchmark corresponding to the current communication terminal from the standard detection policy benchmark library based on the terminal model. The standard detection policy benchmark includes two parts: a standard physical detection trajectory and a standard test parameter configuration. The standard physical detection trajectory is a set of motion paths of the micro-touch probes at the end of the robotic arm, pre-defined for the screen display area of ​​the terminal model. The standard physical detection trajectory consists of a series of ordered three-dimensional coordinate points, each corresponding to a preset touch point on the screen display area. The preset touch points are evenly distributed according to the row and column grid of the screen display area, and the row and column spacing between adjacent preset touch points are preset standard spacing values. The standard test parameter configuration includes standard test procedures for the RF communication function module, the acoustic audio function module, the interface electrical function module, and the screen display function module. Each function module's standard test procedure defines the corresponding test item list, the execution order of each test item, the subsequent robotic arm motion parameters for each test item, and the data acquisition parameters for each test item. The edge nodes sequentially number all three-dimensional coordinate points in the standard physical detection trajectory and arrange the standard test procedures for each function module in the standard test parameter configuration according to the correspondence between detection stations B, C, D, and E, generating full-scale detection execution instructions.

[0042] Therefore, the full-scale detection execution instruction in this embodiment of the invention includes a screen touch detection sub-instruction and test sub-instructions for each functional module. The screen touch detection sub-instruction carries all ordered three-dimensional coordinate points of the standard physical detection trajectory, as well as the standard pressing force and standard pressing duration of the micro touch probe at each preset touch point. The test sub-instructions for each functional module carry the standard test process of the corresponding functional module.

[0043] Step 202: Based on the full-scale detection execution command, control the subsequent robotic arm of the subsequent detection station to perform full-dimensional functional testing on the current communication terminal and obtain functional test data.

[0044] Optionally, the edge node will split the full-scale detection execution instructions according to the detection station type, sending the screen touch detection sub-instruction to the controller of the subsequent robotic arm corresponding to detection station E, the radio frequency communication function module test sub-instruction to the controller of the subsequent robotic arm corresponding to detection station B, the acoustic audio function module test sub-instruction to the controller of the subsequent robotic arm corresponding to detection station C, and the interface electrical function module test sub-instruction to the controller of the subsequent robotic arm corresponding to detection station D.

[0045] Following the ordered sequence of three-dimensional coordinate points in the screen touch detection sub-instruction, the robotic arm of detection station E sequentially drives the end-effector micro-touch probes to move to the three-dimensional coordinate points corresponding to each preset touch point. It applies pressure to the screen display area with a standard pressure level and maintains that pressure for the standard duration. Upon pressing each preset touch point, the screen display area generates a touch response signal. The touch response acquisition unit of detection station E records the touch response time and coordinates corresponding to each preset touch point. The touch response time is the time interval from the application of pressure by the micro-touch probe to the return of a touch response signal from the screen display area, and the touch response coordinates are the actual coordinates of the returned touch response signal from the screen display area. Detection station E reports the touch response times and coordinates corresponding to all preset touch points to the edge node.

[0046] Following the standard test procedure in the test sub-instruction of the RF communication function module, the robotic arm of test station B drives the end-effector micro-touch probe to make physical contact with the RF test contact of the current communication terminal, and sequentially performs signal transmission power test and signal reception sensitivity test in frequency bands F1 to F5. The RF signal analyzer of test station B records the signal transmission power value and signal reception sensitivity value in each frequency band, and reports the corresponding signal transmission power value and signal reception sensitivity value of all frequency bands to the edge node.

[0047] Following the standard test procedure in the acoustic audio function module test sub-instruction, the robotic arm at test station C drives the end effector micro-touch probe to trigger the speaker of the current communication terminal to play a preset standard audio signal. The acoustic sensor at test station C collects the actual audio signal output by the speaker and calculates the audio distortion between the actual audio signal and the standard audio signal. Subsequently, the robotic arm drives the end effector micro-touch probe to trigger the microphone of the current communication terminal to record a preset standard sound source signal. The acoustic sensor at test station C compares the actual recorded signal with the standard sound source signal and calculates the recording signal-to-noise ratio. Test station C then reports the audio distortion and recording signal-to-noise ratio to the edge node.

[0048] Following the standard test procedure in the interface electrical function module test sub-instruction, the subsequent robotic arm of testing station D drives the end effector micro-touch probe to physically contact the charging interface of the current communication terminal, performing a charging interface voltage test. The electrical analyzer at testing station D records the charging interface voltage value. Subsequently, the subsequent robotic arm drives the end effector micro-touch probe to physically contact the data interface of the current communication terminal, performing a data transmission rate test. The electrical analyzer at testing station D records the data transmission rate value. Testing station D then reports the charging interface voltage value and the data transmission rate value to the edge node.

[0049] The edge node receives all the data reported by detection stations B, C, D and E respectively, and summarizes the above data into functional test data.

[0050] Step 203: If the image recognition result indicates the existence of a defect area, then the corresponding position points of the defect area in the terminal coordinate system are determined as spatial defect position points.

[0051] Optionally, the edge node reads the image recognition result. If the image recognition result contains defect contour coordinates, defect area, and defect location identifier, it determines the current communication terminal surface where the defect area is located based on the defect location identifier and retrieves the terminal coordinate system definition file corresponding to the terminal model of the current communication terminal from the server cluster. The terminal coordinate system definition file defines a three-dimensional rectangular coordinate system established with the geometric center of the current communication terminal as the origin. The terminal coordinate system definition file contains the surface equations and surface boundary coordinates of the six surfaces of the current communication terminal in the terminal coordinate system.

[0052] The edge node retrieves the surface equation of the corresponding surface from the terminal coordinate system definition file based on the defect location identifier. It then substitutes the defect contour coordinates into this surface equation for coordinate transformation, converting the defect contour coordinates from the pixel coordinate system of the vision sensor to the terminal coordinate system, thus obtaining the defect contour boundary of the defect region in the terminal coordinate system. The edge node then uniformly samples the region inside the defect contour boundary at a preset sampling interval, obtaining the corresponding location points of the defect region in the terminal coordinate system. These location points are defined as spatial defect location points, each represented by three-dimensional coordinates in the terminal coordinate system.

[0053] Step 204: For spatial traversal type inspection projects, compare the spatial defect location points with the standard physical inspection trajectory in the standard inspection strategy benchmark, and remove non-functional related points in the standard physical inspection trajectory that overlap with the defect area to obtain the target spatial location point. Adjust the detection frequency and detection point density of the related functional points adjacent to the defect area in the target spatial location point, while keeping the standard functional related points unchanged, to generate a differentiated physical inspection trajectory.

[0054] Optionally, the edge nodes extract the three-dimensional coordinates of all preset touch points included in the standard physical inspection trajectory from the standard inspection strategy benchmark. Spatial traversal inspection items refer to inspection items that require the micro-touch probe at the end of the robotic arm to perform touch operations point-by-point on the screen display area according to a preset path. Spatial traversal inspection items correspond to screen touch detection at inspection station E. The Euclidean distance between the three-dimensional coordinates of each preset touch point and all spatial defect location points is calculated one by one. If the Euclidean distance between a preset touch point and any spatial defect location point is less than or equal to a preset overlap judgment threshold, then the preset touch point is determined to overlap with the defect area. If the Euclidean distance between a preset touch point and all spatial defect location points is greater than the overlap judgment threshold, then the preset touch point is determined not to overlap with the defect area.

[0055] Edge nodes remove pre-defined touch points that overlap with the defect area from the standard physical inspection trajectory. Among these pre-defined touch points overlapping with the defect area, it is determined whether the removed touch points are functionally associated points. Functionally associated points refer to points within the pre-defined touch points that correspond to the location of specific functional components within the screen display area. These specific functional components include the front camera opening area, earpiece opening area, and fingerprint recognition area. The three-dimensional coordinates of functionally associated points are pre-marked with functional association identifiers in the standard inspection strategy baseline. If a removed pre-defined touch point has a functional association identifier, then that point is a functionally associated point overlapping with the defect area; that is, functionally associated points among non-functionally associated points are not removed and remain in the trajectory. If a removed pre-defined touch point does not have a functional association identifier, then that point is a non-functionally associated point and is permanently removed from the standard physical inspection trajectory. The remaining pre-defined touch points after the above removal process are determined as target spatial location points.

[0056] Edge nodes identify associated functional points adjacent to the defect area in the target spatial location. The method for determining associated functional points is as follows: a preset touch point with a functional association identifier in the target spatial location point, and the Euclidean distance between the preset touch point and any spatial defect location point is less than or equal to a preset proximity determination threshold, and the proximity determination threshold is greater than the overlap determination threshold.

[0057] For associated function points, the edge node increases their detection frequency from the standard 1 to a preset enhanced frequency, and adds additional detection points around the associated function points at a preset density interval. This density interval is smaller than the standard interval value in the standard physical detection trajectory, thereby increasing the detection point density around the associated function points. Points in the target spatial location point other than the associated function points are standard function points. The detection frequency and density of standard function points remain consistent with the definition in the standard physical detection trajectory and are not adjusted. The edge node rearranges the original 3D coordinates of the standard function points in the target spatial location point, the original 3D coordinates and enhanced detection frequencies of the associated function points, and the 3D coordinates of the additional detection points added around the associated function points into an ordered sequence according to spatial order, generating a differentiated physical detection trajectory.

[0058] Among them, spatial traversal type inspection items refer to inspection items in which the micro-touch probe at the end of the robotic arm needs to move point by point on the screen display area of ​​the communication terminal under test according to a preset spatial path and perform touch operations during the inspection process. Non-functional associated points refer to preset touch points in the standard physical inspection trajectory that do not have functional association indicators. Non-functional associated points do not correspond to the location of specific functional components in the screen display area. Associated functional associated points refer to preset touch points in the target spatial location that have functional association indicators and whose Euclidean distance from the defect area is less than or equal to the proximity judgment threshold. Standard functional associated points refer to the remaining points in the target spatial location that are not associated functional associated points, including functional associated points that are far away from the defect area and all non-functional associated reserved points. Inspection frequency refers to the number of times the micro-touch probe at the end of the robotic arm performs touch operations at the same inspection point. Inspection point density refers to the number of inspection points distributed in a unit area around the inspection point.

[0059] Step 205: For module-independent inspection items, match the corresponding target test parameter configuration in the standard test parameter configuration in the standard inspection strategy benchmark based on the spatial proximity between the defect area and each functional module node.

[0060] Optionally, module-independent testing projects refer to testing projects that do not require a robotic arm to perform spatial traversal touch operations on the screen display area, but instead perform specialized functional tests on each independent functional module of the current communication terminal. Module-independent testing projects include the radio frequency communication function test of testing station B, the acoustic audio function test of testing station C, and the interface electrical function test of testing station D. The edge node retrieves the terminal coordinate system definition file corresponding to the terminal model of the current communication terminal from the server cluster. The terminal coordinate system definition file also contains the three-dimensional coordinates of each functional module node in the terminal coordinate system. The functional module nodes include the radio frequency communication module node, the acoustic audio module node, and the interface electrical module node. The edge node calculates the Euclidean distance between the geometric center point of the defect area and the three-dimensional coordinates of each functional module node, and determines the calculated Euclidean distance as the spatial proximity between the defect area and each functional module node. The geometric center point of the defect area is the arithmetic mean of the three-dimensional coordinates of all spatial defect location points.

[0061] The standard test parameter configuration in the standard testing strategy benchmark is divided into multiple parameter configuration levels based on different spatial proximity ranges. Each parameter configuration level corresponds to a proximity interval and a set of test parameter configuration schemes. Edge nodes match the calculated spatial proximity of each functional module node with the proximity intervals of each parameter configuration level in the standard test parameter configuration to determine the proximity interval into which each functional module node falls. The test parameter configuration scheme corresponding to this proximity interval is then determined as the target test parameter configuration for that functional module node. If the spatial proximity of a functional module node falls into the nearest neighbor interval, the target test parameter configuration for that functional module node increases the test sampling accuracy and the number of test items based on the standard test procedure. If the spatial proximity of a functional module node falls into the middle neighbor interval, the target test parameter configuration for that functional module node increases the test sampling accuracy while maintaining the same number of test items based on the standard test procedure. If the spatial proximity of a functional module node falls into the farthest neighbor interval, the target test parameter configuration for that functional module node is completely consistent with the standard test procedure.

[0062] Among them, module-independent testing items refer to testing items that perform specialized functional tests on each independent functional module of the current communication terminal, without requiring spatial traversal touch operations on the screen display area. A functional module node refers to the geometric center position of each functional module within the current communication terminal in the terminal coordinate system, represented by three-dimensional coordinates in the terminal coordinate system. Spatial proximity refers to the Euclidean distance between the geometric center point of the defect area and the three-dimensional coordinates of the functional module node, measured in millimeters. Target test parameter configuration refers to the test parameter scheme obtained based on spatial proximity, used to replace the standard test procedure for the corresponding functional module in the standard test parameter configuration.

[0063] Step 206: Based on the differentiated physical detection trajectory and target test parameter configuration, generate optimized detection execution instructions, and control the subsequent robotic arm of the subsequent detection station to perform full-dimensional functional tests on the current communication terminal based on the optimized detection execution instructions, and obtain functional test data.

[0064] Optionally, the edge node orchestrates the differentiated physical detection trajectory and target test parameter configuration to generate optimized detection execution instructions. These optimized instructions include optimized screen touch detection sub-instructions and optimized test sub-instructions for each functional module. The optimized screen touch detection sub-instruction carries an ordered three-dimensional coordinate sequence of all points in the differentiated physical detection trajectory, the detection frequency for each point, and the standard pressing pressure and duration of the micro-touch probe at each point. Each functional module's optimized test sub-instruction carries the target test parameter configuration for the corresponding functional module. The edge node sends the optimized screen touch detection sub-instructions to the controller of the subsequent robotic arm corresponding to detection station E, the optimized test sub-instructions for the RF communication functional module to the controller of the subsequent robotic arm corresponding to detection station B, the optimized test sub-instructions for the acoustic audio functional module to the controller of the subsequent robotic arm corresponding to detection station C, and the optimized test sub-instructions for the interface electrical functional module to the controller of the subsequent robotic arm corresponding to detection station D.

[0065] Following the ordered three-dimensional coordinate sequence in the optimized screen touch detection sub-instruction, the robotic arm of detection station E sequentially drives the end-effector micro-touch probe to each location. It applies pressure to the screen display area with a standard pressure intensity and duration according to the detection frequency corresponding to that location. The touch response collector records the touch response time and coordinates for each press. For associated function locations with a detection frequency greater than once, the touch response collector records the touch response time and coordinates for each press. The arithmetic mean of the touch response times from multiple presses is taken as the final touch response time for that location, and the coordinate that appears most frequently among the multiple presses is taken as the final touch response coordinate for that location. Detection station E reports the final touch response times and coordinates for all locations to the edge node.

[0066] The subsequent robotic arm of testing station B executes test actions according to the target test parameters configured in the optimized test sub-instruction of the radio frequency communication function module. If the target test parameter configuration adds an additional frequency band and an increased sampling rate compared to the standard test procedure, the subsequent robotic arm performs signal transmission power test and signal reception sensitivity test at the increased sampling rate in both the standard frequency band and the additional frequency band. The radio frequency signal analyzer of testing station B records the signal transmission power value and signal reception sensitivity value collected at the increased sampling rate in each frequency band and reports them to the edge node.

[0067] The robotic arm of detection station C executes test actions according to the target test parameters configured in the optimized test sub-instruction of the acoustic audio function module. The acoustic sensors of detection station C collect audio distortion and recording signal-to-noise ratio and report them to the edge node.

[0068] The subsequent robotic arm of testing station D executes test actions according to the target test parameters configured in the optimized test sub-instruction of the interface electrical function module. The electrical analyzer of testing station D records the charging interface voltage value and data transmission rate value and reports them to the edge node.

[0069] The edge node receives all the data reported by detection stations B, C, D and E respectively, and summarizes the above data into functional test data.

[0070] The embodiments of the present invention enhance the detection accuracy of key areas by reducing invalid detection points through a differentiated detection strategy, so that the functional test data can accurately reflect the actual impact of defective areas on the functional modules of the communication terminal, thereby improving the detection accuracy and price evaluation accuracy of the communication terminal testing process.

[0071] Optionally, the processes of steps 301 to 306 include: Step 301: If the image recognition result indicates that there is no defective area, then the performance test indicators of each test point in the functional test data are mapped to the corresponding internal component nodes in the three-dimensional digital model.

[0072] Optionally, the edge node reads the judgment identifier from the image recognition result. If the judgment identifier indicates no defects, the 3D digital model corresponding to the current communication terminal model is retrieved from the server cluster. In this embodiment of the invention, the 3D digital model includes the outer shell geometry of the current communication terminal and the 3D spatial positions of each internal functional component. The geometric center position of each functional component is marked in the form of internal component nodes in the 3D digital model. Each internal component node has a unique component identifier and 3D coordinates in the 3D digital model coordinate system. Functional components include, but are not limited to, a screen touch layer, an RF communication antenna, a speaker unit, a microphone unit, a charging interface module, a data transmission interface module, and a display panel.

[0073] Edge nodes extract performance metrics for each test point from the functional test data. Performance metrics refer to the numerical test results of each test item in the functional test data. Based on the preset correspondence between each test item and functional component in the functional test data, the performance metrics of each test item are associated with the corresponding internal component node. The preset correspondence is stored in the server cluster. The preset correspondence defines the internal component node of the screen touch layer for each preset touch point of test station E (touch response time and touch response coordinates), the internal component node of the RF communication antenna for each frequency band signal transmit power and signal receive sensitivity for test station B, the internal component node of the speaker unit for the audio distortion of test station C, the internal component node of the microphone unit for the recording signal-to-noise ratio of test station C, the internal component node of the charging interface module for the charging interface voltage value of test station D, the internal component node of the data transmission rate module for test station D, and the internal component node of the display panel for the brightness uniformity test results and color accuracy deviation test results of test station E.

[0074] The edge node writes the component identifier, 3D coordinates, and associated performance test indicators of each internal component node into the corresponding internal component node data structure in the 3D digital model, thus completing the mapping of functional test data to the 3D digital model.

[0075] Internal component nodes refer to the geometric center points of each functional component marked in the 3D digital model. Each internal component node has a unique component identifier and 3D coordinates in the 3D digital model coordinate system. Performance test indicators refer to the numerical test results produced by each test item in the functional test data. Performance test indicators include touch response time, touch response coordinates, signal transmission power value, signal reception sensitivity value, audio distortion, recording signal-to-noise ratio, charging interface voltage value, data transmission rate value, brightness uniformity test results, and color accuracy deviation test results.

[0076] Step 302: Based on the spatial adjacency relationship between internal component nodes in the three-dimensional digital model, analyze the spatial dispersion of performance detection indicators of adjacent detection points to obtain the performance degradation region.

[0077] Optionally, the edge node extracts the 3D coordinates of all internal component nodes from the 3D digital model and constructs an adjacency table based on the spatial adjacency relationships between the internal component nodes. Spatial adjacency refers to the condition that two internal component nodes have a spatial adjacency relationship when the Euclidean distance between them is less than or equal to a preset adjacency distance threshold. The Euclidean distance between any two internal component nodes is calculated one by one, and pairs of internal component nodes whose Euclidean distance is less than or equal to the adjacency distance threshold are recorded in the adjacency table. The adjacency table records internal component nodes with spatial adjacency relationships in the form of node pairs.

[0078] For each pair of spatially adjacent internal component nodes in the adjacency table, the edge node reads the associated performance metrics of each pair from the 3D digital model and calculates the deviation between these metrics. The deviation is calculated as follows: for the same type of performance metric, the absolute value of the difference between the same type of performance metric for two internal component nodes is calculated, and this absolute value is divided by the interval width of the standard parameter range corresponding to that type of performance metric to obtain the normalized deviation value. If there are multiple types of performance metric values ​​between a pair of internal component nodes, the arithmetic mean of the normalized deviation values ​​for each type of performance metric is calculated to obtain the comprehensive deviation value for the pair of internal component nodes.

[0079] The edge node compares the deviation value of the comprehensive index with a preset dispersion threshold. If the deviation value of the comprehensive index is greater than the dispersion threshold, it is determined that there is spatial dispersion in the performance detection index between the pair of internal component nodes, and the edge node marks the pair of internal component nodes as a performance discrete node pair.

[0080] Edge nodes cluster the internal component nodes involved in all discrete performance node pairs. The clustering rule is: if two discrete performance node pairs have common internal component nodes, then the two discrete performance node pairs are assigned to the same cluster group. The spatial region enclosed by the internal component nodes in each cluster group in the 3D digital model is defined as a performance degradation region. Edge nodes record the list of internal component nodes contained in each performance degradation region and the regional location of the performance degradation region. The regional location is the arithmetic mean of the 3D coordinates of all internal component nodes within the performance degradation region.

[0081] Spatial adjacency refers to the spatial relationship between two internal component nodes when the Euclidean distance between them is less than or equal to the adjacency distance threshold. The adjacency distance threshold is a pre-set Euclidean distance threshold used to determine whether two internal component nodes have a spatial adjacency relationship; the unit of the adjacency distance threshold is millimeters. The index deviation value refers to the degree of difference in the same type of performance test index between a pair of spatially adjacent internal component nodes, expressed as a normalized numerical value.

[0082] Spatial dispersion refers to the degree of performance difference reflected by the comprehensive index deviation between a pair of spatially adjacent internal component nodes. Performance degradation region refers to the spatial region enclosed by clustered performance-discrete node pairs, characterizing the existence of performance anomalies within the current communication terminal.

[0083] Step 303: Based on the number of performance degradation regions and the impact of the location of each performance degradation region on the performance of the communication terminal, generate a full-dimensional state mapping result.

[0084] Optionally, the edge node counts the number of performance degradation regions and, starting from the location of each performance degradation region, analyzes the impact of each performance degradation region on the overall performance of the current communication terminal.

[0085] Edge nodes retrieve the performance impact weight table corresponding to the terminal model of the current communication terminal from the server cluster. The performance impact weight table defines the performance impact weight coefficients for different spatial location intervals in the 3D digital model coordinate system. These spatial location intervals are divided according to the functional importance within the current communication terminal; the higher the functional importance, the larger the performance impact weight coefficient. The edge nodes match the location of each performance degradation region with the spatial location intervals in the performance impact weight table to determine which interval each performance degradation region falls into, and then define the performance impact weight coefficient corresponding to that interval as the performance impact weight coefficient for that performance degradation region.

[0086] The edge nodes sum the performance impact weighting coefficients for all performance degradation regions to obtain a comprehensive performance impact index. The edge nodes retrieve the status level classification rules corresponding to the current communication terminal model from the server cluster. These rules define the terminal status levels corresponding to different comprehensive performance impact index intervals, with four levels: Excellent, Good, Average, and Poor. The edge nodes match the comprehensive performance impact index with the status level classification rules to determine the current communication terminal's status level.

[0087] Edge nodes encapsulate the number of performance degradation regions, the location of each performance degradation region, the list of internal component nodes contained in each performance degradation region, the performance impact weight coefficient of each performance degradation region, the comprehensive performance impact index, and the terminal state level into a full-dimensional state mapping result.

[0088] Step 304: If the image recognition result indicates the existence of a defective area, map the defective area to the surface mesh of the three-dimensional digital model, and associate the functional test data with the internal component nodes of the three-dimensional digital model.

[0089] Optionally, the edge node reads the image recognition result. If the image recognition result contains defect contour coordinates, defect area, and defect location identifier, it retrieves the 3D digital model corresponding to the current communication terminal model from the server cluster. Optionally, in this embodiment of the invention, the outer shell geometry of the 3D digital model is expressed using a triangular surface mesh. The surface mesh is composed of a large number of triangular facets, and each triangular facet has three vertices with 3D coordinates in the 3D digital model coordinate system and a surface identifier of the surface to which the triangular facet belongs.

[0090] All spatial defect locations are projected onto the surface mesh of the 3D digital model. The projection method is as follows: for each spatial defect location, the nearest triangular facet to that location is found in the surface mesh, and the spatial defect location is marked on that facet. All marked triangular facets are defined as defect mapping facets. The area enclosed by all defect mapping facets on the surface mesh of the 3D digital model is the mapping representation of the defect region in the 3D digital model. A list of facet identifiers for all defect mapping facets is recorded, and the defect area and defect location identifier are written into the defect mapping facet data structure.

[0091] The edge node extracts the performance test indicators of each test item from the functional test data, and associates the performance test indicators of each test item with the corresponding internal component node in the three-dimensional digital model according to the preset correspondence in step 301, and writes the performance test indicators into the data structure of the corresponding internal component node.

[0092] In this context, the surface mesh refers to the set of triangular facets in the 3D digital model used to represent the geometric surface of the current communication terminal's outer shell; the surface mesh is composed of a large number of triangular facets. The defect mapping facet refers to the triangular facet in the surface mesh marked with spatial defect location points; the defect mapping facet is used to represent the spatial distribution of defect regions in the 3D digital model.

[0093] Step 305: Based on the spatial topological relationship between the defect area and the internal component nodes in the three-dimensional digital model, determine the abnormal impact of the defect area on the internal components.

[0094] Optionally, the edge nodes determine the abnormal impact of the defective region on the internal components based on the spatial topological relationship between the defective region and the internal component nodes in the three-dimensional digital model, as in steps 3051 to 3054.

[0095] Step 306: Based on the results of the anomaly impact and the functional test data, generate a full-dimensional state mapping result.

[0096] Optionally, the edge node extracts the component identifier, affected type, specific judgment category, and test result status list for each affected component from the abnormal impact results. Affected types include physical damage, indirect stress impact, and electrical conduction impact. Specific judgment categories include functional failure nodes, functionally limited nodes, stress-induced failure nodes, non-stress-related failure nodes, cascading failure nodes, and non-cascading failure nodes. The edge node extracts performance test indicators associated with all internal component nodes from the functional test data. Performance test indicators refer to the numerical test results generated by each test item in the functional test data, including touch response time, touch response coordinates, signal transmission power value, signal reception sensitivity value, audio distortion, recording signal-to-noise ratio, charging interface voltage value, data transmission rate value, brightness uniformity test results, and color accuracy deviation test results.

[0097] Edge nodes categorize and integrate all internal component nodes in the 3D digital model according to their existence status in the anomaly impact results. The categorization and integration rules are as follows: if the component identifier of an internal component node exists in the anomaly impact results, then the internal component node is marked as an anomaly impact node, and the affected type, specific judgment category, and test result status list corresponding to the component identifier in the anomaly impact results, as well as the performance test indicators associated with the internal component node in the functional test data, are written into the data structure of the internal component node in the 3D digital model.

[0098] If the component identifier of a certain internal component node does not exist in the abnormal impact results, the internal component node is marked as an unaffected node, and the performance test indicators and corresponding test result status associated with the internal component node in the functional test data are written into the data structure of the internal component node in the 3D digital model.

[0099] The edge node obtains the list of patch identifiers, defect area, and defect location identifiers of the defect-mapped patches from step 304. The edge node combines the list of patch identifiers, defect area, and defect location identifiers of the defect-mapped patches with the categorized and integrated data structure of all internal component nodes (including the affected type, specific judgment category, test result status list, and performance test indicators for nodes affected by anomalies, as well as the performance test indicators and test result status for unaffected nodes) to obtain the full-dimensional state mapping result. The full-dimensional state mapping result is output in the form of a structured data file. The structured data file contains a surface defect mapping data segment and an internal component state data segment. The surface defect mapping data segment records relevant information about the defect-mapped patches, and the internal component state data segment records the state information of all internal component nodes.

[0100] An anomalously affected node refers to an internal component node whose component identifier exists in the anomalous impact results. Anomalously affected nodes include those with physical damage, indirect stress waves, and electrical conducted waves. An unaffected node refers to an internal component node whose component identifier does not exist in the anomalous impact results. An unaffected node indicates that the internal component node is not affected by the anomalous impact of the defect area in terms of spatial topology and electrical connection.

[0101] The surface defect mapping data segment refers to the data segment in the full-dimensional state mapping result used to record the list of facet identifiers, defect area, and defect location identifiers for the defect-mapped facets. The internal component state data segment refers to the data segment in the full-dimensional state mapping result used to record the component identifiers, 3D coordinates, affected state markers, and their associated performance testing indicators and test result states for all internal component nodes.

[0102] This invention implements a full-dimensional state association and evaluation from surface to interior, from local to global, and from physical space to electrical topology through two state mapping paths for both defect-free and defective scenarios. This enables the final full-dimensional state mapping result to comprehensively and accurately reflect the actual impact of surface defects on internal functional components and the classification of abnormal causes, thereby improving the detection accuracy and price evaluation accuracy of the communication terminal testing process.

[0103] Optionally, the processes of steps 3051 to 3054 include: Step 3051: Construct a minimum axis-aligned bounding box based on the coordinates of each spatial defect location point in the coordinate system of the three-dimensional digital model to obtain the defect region bounding box.

[0104] Optionally, the edge node extracts the three-dimensional coordinates of each spatial defect location point in the terminal coordinate system from all spatial defect location points. The three-dimensional coordinates of each spatial defect location point include x-axis coordinates, y-axis coordinates, and z-axis coordinates. The edge node selects the minimum and maximum x-axis coordinates from all spatial defect location points, the minimum and maximum y-axis coordinates from all spatial defect location points, and the minimum and maximum z-axis coordinates from all spatial defect location points.

[0105] Edge nodes determine the bounding box's boundary range along the x-axis using the minimum and maximum x-axis coordinate values, along the y-axis using the minimum and maximum y-axis coordinate values, and along the z-axis using the minimum and maximum z-axis coordinate values. The cuboid bounded by these three axis-aligned boundary ranges is the minimum axis-aligned bounding box. The minimum axis-aligned bounding box is the smallest cuboid whose faces are parallel to the coordinate axes of the terminal coordinate system and can contain all spatial defect locations. Edge nodes define the minimum axis-aligned bounding box as the defect region bounding box and record its six boundary values: minimum x-axis value, maximum x-axis value, minimum y-axis value, maximum y-axis value, minimum z-axis value, and maximum z-axis value.

[0106] The minimum axis-aligned bounding box refers to the smallest cuboid whose faces are parallel to the coordinate axes of the terminal coordinate system and can contain all specified points. The minimum axis-aligned bounding box is determined by six boundary values. The defect region bounding box refers to the minimum axis-aligned bounding box constructed from the coordinates of all spatial defect location points. The defect region bounding box is used to represent the spatial extent occupied by the defect region in the three-dimensional digital model.

[0107] Step 3052: Perform intersection detection based on the geometric properties of the bounding box of the defect region and each internal component node in the 3D digital model to obtain the intersection detection result.

[0108] Optionally, the edge node retrieves the geometric attributes of each internal component node in the 3D digital model corresponding to the current communication terminal model from the server cluster. The geometric attributes include the bounding box of the internal component node and a pre-stored stress transmission distance threshold. The bounding box of an internal component node is a predefined cuboid that encloses the space occupied by the functional component entity corresponding to that internal component node in the 3D digital model. Each face of the bounding box is parallel to the coordinate axes of the terminal coordinate system. The bounding box is determined by six boundary values, which are the minimum and maximum values ​​of the functional component entity corresponding to that internal component node in the x-axis, y-axis, and z-axis directions, respectively. The pre-stored stress transmission distance threshold is a distance value pre-set based on the structural fragility of the functional component. The pre-stored stress transmission distance threshold characterizes the maximum transmission distance that the functional component can withstand without functional abnormality when subjected to external stress, and is expressed in millimeters.

[0109] For each internal component node in the 3D digital model, the edge node performs intersection detection between the bounding box of the defect region and the geometric bounding box of that internal component node. The intersection detection criteria are as follows: If the boundary ranges of the defect region bounding box and the geometric bounding box of the internal component node overlap in the x-axis direction, the boundary ranges of the defect region bounding box overlap in the y-axis direction, and the boundary ranges of the internal component node overlap in the z-axis direction, then the defect region bounding box and the geometric bounding box of the internal component node are determined to have volumetric overlap. If the boundary ranges of the defect region bounding box and the geometric bounding box of the internal component node in exactly one of the three axes have only equal endpoints and no interval overlap, and the boundary ranges in the other two axes overlap, then the defect region bounding box and the geometric bounding box of the internal component node are determined to have surface contact. If the boundary ranges of the defect region bounding box and the geometric bounding box of the internal component node in at least one of the three axes have no overlap and the endpoints are not equal, then the defect region bounding box and the geometric bounding box of the internal component node are determined to have no overlap.

[0110] The criteria for determining overlapping boundary ranges along the x-axis are: the maximum x-axis value of the bounding box of the defect region is greater than the minimum x-axis value of the bounding box of the internal component node, and the minimum x-axis value of the bounding box of the defect region is less than the maximum x-axis value of the bounding box of the internal component node. The criteria for overlapping boundary ranges along the y-axis and z-axis are similar to those for the x-axis; simply substitute the minimum and maximum values ​​of the corresponding axes into the criteria. Edge nodes record the intersection detection results for each internal component node as intersection detection results. In the intersection detection results, the detection result for each internal component node is one of three: volume overlap, surface contact, or no overlap.

[0111] Step 3053: If the intersection detection result indicates that the bounding box of the internal component node has volume overlap or surface contact with the bounding box of the defect area, then the internal component node is identified as a physically damaged node. If the intersection detection result indicates that the bounding box of the internal component node does not overlap with the bounding box of the defect area, then the distance between the center coordinates of the internal component node and the nearest boundary of the bounding box of the defect area is calculated, and internal component nodes whose distance is less than the pre-stored stress transmission distance threshold are identified as indirect stress-affected nodes.

[0112] Optionally, the edge nodes read the intersection detection results of each internal component node one by one. If the intersection detection result of an internal component node is volume overlap or surface contact, then the internal component node is directly identified as a physically damaged node. A physically damaged node indicates that the functional component entity corresponding to the internal component node has a direct physical overlap relationship with the defect area in three-dimensional space, and the surface physical damage of the defect area may directly affect the functional component. If the intersection detection result of an internal component node is no overlap, then the distance between the center coordinates of the internal component node and the nearest boundary of the defect area bounding box is calculated. The center coordinates of the internal component node are the three-dimensional coordinates of the internal component node in the terminal coordinate system. The nearest boundary of the defect area bounding box refers to the boundary face on the six boundary faces of the defect area bounding box that is closest to the center coordinates of the internal component node.

[0113] The method for calculating the distance between the center coordinates of the internal component node and the nearest boundary of the defect region's bounding box is as follows: Calculate the shortest distance from the center coordinates of the internal component node to the corresponding boundary of the defect region's bounding box along the x, y, and z axes. In the x-axis direction, if the x-coordinate of the internal component node's center is less than the minimum x-coordinate of the defect region's bounding box, the shortest distance in the x-axis direction is the minimum x-coordinate of the defect region's bounding box minus the x-coordinate of the internal component node's center. If the x-coordinate of the internal component node's center is greater than the maximum x-coordinate of the defect region's bounding box, the shortest distance in the x-axis direction is the same as the minimum x-coordinate of the internal component node's center. If the x-coordinate of the internal component node's center is between the minimum and maximum x-coordinate of the defect region's bounding box, the shortest distance in the x-axis direction is 0. The calculation methods for the shortest distances in the y and z axes are similar to those for the x-axis direction. The edge node calculates the square root of the sum of the squares of the shortest distances in the x-axis, y-axis, and z-axis directions, and then obtains the distance between the center coordinates of the internal component node and the nearest boundary of the bounding box of the defect region.

[0114] The edge node compares the obtained distance with the pre-stored stress transmission distance threshold corresponding to the internal component node. If the distance is less than the pre-stored stress transmission distance threshold, the internal component node is identified as an indirect stress-affected node. An indirect stress-affected node indicates that although the functional component entity corresponding to the internal component node does not have a direct physical overlap with the defect area in three-dimensional space, the stress generated in the defect area may be transmitted to the functional component through the structural medium due to its proximity to the defect area. If the distance is greater than or equal to the pre-stored stress transmission distance threshold, the internal component node is neither a physically damaged node nor an indirect stress-affected node, and the edge node marks the internal component node as an unaffected node.

[0115] Step 3054: Based on the physical damage nodes, indirect stress wave nodes, and the electrical connection relationships of internal component nodes in the three-dimensional digital model, the abnormal impact results are obtained by traversing the model.

[0116] Optionally, the edge node retrieves the electrical connection relationship between the internal component nodes corresponding to the terminal model of the current communication terminal from the server cluster. Based on the physically damaged nodes, the indirect stress wave and the nodes, and the electrical connection relationship of the internal component nodes in the three-dimensional digital model, the edge node traverses the process to obtain the abnormal impact result, as in steps 30541 to 30544.

[0117] The embodiments of the present invention take into account the physical overlap, stress transmission and electrical transmission effects of defective areas on internal components, making the results of abnormal effects more comprehensive and accurate, and improving the detection accuracy of the communication terminal testing process and the accuracy of price evaluation.

[0118] Optionally, the processes of steps 30541 to 30544 include: Step 30541: Based on the electrical connection relationship, traverse the system starting from each physically damaged node, retrieving first-level associated nodes directly electrically connected to the starting node, and continuing to retrieve second-level associated nodes directly electrically connected to the first-level associated nodes, until the preset maximum number of electrical cascading layers is reached, thus obtaining the electrical conduction affected nodes. Direct electrical connection is characterized by direct connection through wires, ribbon cables, or printed circuit boards.

[0119] Optionally, the electrical connection relationships in this embodiment of the invention are stored in the form of a directed graph. Nodes in the directed graph are component identifiers of internal component nodes. Directed edges in the directed graph represent direct electrical connections between two internal component nodes established through wires, ribbon cables, or printed circuit boards. The direction of the directed edge is from the signal or power provider to the signal or power receiver. Each directed edge is accompanied by a connection type identifier, which includes power connection, data signal connection, and control signal connection.

[0120] Direct electrical connection indicates that two internal component nodes are directly connected via wires, ribbon cables, or printed circuit boards without passing through other internal component nodes.

[0121] Edge nodes start with the component identifier of each physically damaged node and perform a level-by-level traversal of the directed graph of electrical connections, following the following traversal rules: Starting from the starting node, internal component nodes directly electrically connected to the starting node are retrieved along directed edges. These retrieved internal component nodes are identified as first-level associated nodes, which are internal component nodes with a direct electrical connection to the physically damaged node. Starting from each first-level associated node, internal component nodes directly electrically connected to the first-level associated node are retrieved along directed edges. These retrieved internal component nodes are identified as second-level associated nodes, which are internal component nodes with an electrical connection of one level to the physically damaged node. This process continues, retrieving the Nth-level associated node from the N-1th-level associated node, until the traversal reaches a preset maximum electrical cascading layer. The preset maximum electrical cascading layer is a pre-set upper limit value for traversal depth, used to limit the maximum number of layers traversed from the starting node. During the traversal, component identifiers already identified as physically damaged nodes or indirectly affected by stress waves are excluded to avoid duplicate marking. Edge nodes, after excluding physically damaged nodes and indirectly affected by stress waves from all the first-level, second-level, and up to the Nth-level associated nodes obtained through traversal, identify the remaining internal component nodes as electrically conducted nodes.

[0122] Step 30542: Match the functional module corresponding to each test point in the functional test data with the functional module associated with each internal component node in the 3D digital model to obtain the target component node, and associate each test point with its corresponding target component node to obtain the component function association relationship.

[0123] Optionally, the edge node extracts a list of test points corresponding to each test item and the functional module identifier tested at each test point from the functional test data. A test point refers to the physical test location or logical test item unit when the robotic arm performs a specific test action during the full-dimensional functional test. The functional module identifier indicates the type of functional component tested at the test point, including screen touch layer, RF communication antenna, speaker unit, microphone unit, charging interface module, data transmission interface module, and display panel. The edge node extracts the component identifier of each internal component node and the corresponding functional module identifier from the 3D digital model. It compares the functional module identifier of each test point in the functional test data with the functional module identifier of each internal component node in the 3D digital model one by one. If the functional module identifier of a certain test point matches the functional module identifier of a certain internal component node, then that internal component node is determined as the target component node for that test point. The edge node establishes a mapping relationship between the point identifier of each test point and the component identifier of its corresponding target component node, and summarizes the mapping relationship between all test points and their target component nodes into a component functional association relationship.

[0124] Step 30543: Based on the component functional association relationship, assign the test result status of each detection point to its corresponding target component node to obtain the status mapping relationship.

[0125] Optionally, the edge node extracts the test result status of each test point from the functional test data.

[0126] The test result status is a status indicator obtained by comparing the performance test index of the test point with the corresponding standard parameter range. The test result status includes success and failure. The test result status of the test point whose performance test index is within the corresponding standard parameter range is success. The test result status of the test point whose performance test index exceeds the corresponding standard parameter range is failure. The standard parameter range is predefined in the standard test strategy benchmark.

[0127] Edge nodes assign the test result status of each detection point to the corresponding target component node according to the component functional association. If a target component node corresponds to only one detection point, then the test result status of the target component node is the test result status of that detection point. If a target component node corresponds to multiple detection points, then the test result status of all detection points corresponding to the target component node is retained, and all test result statuses are recorded in the status field of the target component node in list form. All target component nodes and their corresponding test result status lists are summarized into a status mapping relationship.

[0128] Step 30544: Based on the electrical connection relationship, state mapping relationship, physical damaged nodes, indirect stress wave nodes, and electrical conduction wave nodes, determine the abnormal impact results.

[0129] Optionally, the edge nodes determine the abnormal impact results based on electrical connection relationships, state mapping relationships, physically damaged nodes, indirect stress-affected nodes, and electrically conducted-affected nodes, as in steps 305441 to 305444.

[0130] This invention, through damage function determination, stress function determination, and electrical cascading function determination, accurately distinguishes the abnormal states of functional failure nodes, functionally limited nodes, stress-induced failure nodes, non-stress-related failure nodes, cascading failure nodes, and non-cascading failure nodes. This ensures that the results of abnormal impacts accurately reflect whether the functional abnormality of each affected internal component is directly caused by physical damage, indirectly caused by stress transmission, or caused by electrical cascading transmission. This provides a basis for the abnormal impacts in subsequent recycling price assessments, improving the detection accuracy and price assessment accuracy of the communication terminal testing process.

[0131] Optionally, the process of steps 305441 to 305444 includes: Step 305441: Perform state matching based on the state mapping relationship. If the test result of the physically damaged node is all failures, then the physically damaged node is determined as a functional failure node; otherwise, the physically damaged node is determined as a functionally limited node, and the damaged function determination result is obtained.

[0132] Optionally, the edge node reads the test result status list corresponding to each physically damaged node one by one from the status mapping relationship, and checks whether all test result statuses in the test result status list of the physically damaged node are failures. If all test result statuses of all detection points corresponding to the physically damaged node are failures, it indicates that the functional component corresponding to the physically damaged node has completely lost its function, and the physically damaged node is identified as a functional failure node. If the test result status list corresponding to the physically damaged node contains both successes and failures, or all successes, it indicates that the functional component corresponding to the physically damaged node still retains some or all of its functions, but may be at risk of performance degradation due to physical damage, and the physically damaged node is identified as a functionally restricted node. The edge node summarizes the component identifier of each physically damaged node and the corresponding functional failure node or functionally restricted node determination results into a damaged function determination result.

[0133] Step 305442: Perform state matching based on the state mapping relationship. If the test result of the indirect stress-affected node is a failure and the electrical connection relationship indicates that there is at least one directly electrically connected physically damaged node, then the indirect stress-affected node is identified as a stress-induced failure node. Otherwise, the indirect stress-affected node is identified as a non-stress-related fault node, and the stress function determination result is obtained.

[0134] Optionally, the edge node reads the test result status list corresponding to each indirect stress-affected node one by one from the state mapping relationship, and checks whether there is at least one failed test result status in the test result status list of the indirect stress-affected node. If all test result statuses in the test result status list of the indirect stress-affected node are successful, then the indirect stress-affected node is directly identified as a non-stress-related fault node. A non-stress-affected fault node indicates that although the indirect stress-affected node is spatially within the stress transmission range, no abnormality was found in the functional test. If there is at least one failed test result status in the test result status list of the indirect stress-affected node, then the electrical connection relationship is searched to see if the indirect stress-affected node is directly electrically connected to at least one physically damaged node. The search method is as follows: search all directed edges with the indirect stress-affected node as the endpoint in the directed graph of the electrical connection relationship, and determine whether there is a component identifier of at least one physically damaged node at the other endpoint of these directed edges. If there is a component identifier of at least one physically damaged node, it indicates that the functional abnormality of the indirect stress-affected node may be caused by the stress transmitted by the physically damaged node through the electrical path, and the indirect stress-affected node is identified as a stress-induced failure node. If there is no component identifier for a physically damaged node, it indicates that there is no direct electrical conduction path between the functional abnormality of the indirect stress-affected node and the physically damaged node, and the indirect stress-affected node is identified as a non-stress-related fault node.

[0135] The edge node summarizes the component identification of each indirect stress-affected node and the determination results of its corresponding stress-induced failure node or non-stress-related fault node into a stress function determination result.

[0136] Step 305443: Perform state matching based on the state mapping relationship. If the test result of the electrical conduction wave node is a failure, and the electrical connection relationship indicates that the upstream dependent node of the electrical conduction wave node is a physically damaged node or the upstream dependent node includes at least one functionally failed node, then the electrical conduction wave node is determined as a cascading failure node. Otherwise, the electrical conduction wave node is determined as a non-cascading failure node, and the electrical cascading function determination result is obtained.

[0137] Optionally, the edge node reads the test result status list corresponding to each electrically conducted sweep node one by one from the state mapping relationship, and checks whether there is at least one failed test result status in the test result status list of the electrically conducted sweep node. If all the test result status lists of the electrically conducted sweep node are successful, then the electrically conducted sweep node is directly identified as a non-chain failure node. A non-chain failure node means that although the electrically conducted sweep node is located downstream of the physically damaged node in terms of electrical connection relationship, no abnormality was found in the functional test.

[0138] If at least one test result status is failed in the test result status list of an electrical conducted sweep node, then the upstream dependent node of that electrical conducted sweep node is determined in the electrical connection relationship. The upstream dependent node is determined as follows: in the directed graph of the electrical connection relationship, find all directed edges whose endpoints are the electrical conducted sweep node. The internal component nodes corresponding to the starting points of these directed edges are the upstream dependent nodes of that electrical conducted sweep node. The upstream dependent node is the source node that provides power or data signals to the electrical conducted sweep node in the electrical connection relationship.

[0139] The edge node checks all upstream dependent nodes for component identifiers belonging to physically damaged nodes or functionally failed nodes. The functionally failed node is the determination result identified in step 305441. If at least one physically damaged node or at least one functionally failed node exists among the upstream dependent nodes, it indicates that the functional abnormality of the electrical conduction affected node may be caused by the cascading propagation of upstream physical damage or functional failure along the electrical path, and the electrical conduction affected node is identified as a cascading failure node. If neither physically damaged nor functionally failed nodes exist among the upstream dependent nodes, it indicates that there is no cascading propagation relationship between the functional abnormality of the electrical conduction affected node and the physical damage or functional failure of the upstream nodes, and the electrical conduction affected node is identified as a non-cascading failure node. The edge node summarizes the component identifier of each electrical conduction affected node and the corresponding cascading or non-cascading failure node determination results into an electrical cascading function determination result.

[0140] Step 305444: Based on the results of the damage function assessment, stress function assessment, and electrical interlock function assessment, the abnormal impact results are combined to obtain the results.

[0141] Optionally, the edge node extracts the component identifier of each physically damaged node and the corresponding functional failure node or functionally limited node from the damaged function judgment result; extracts the component identifier of each indirect stress-affected node and the corresponding stress-induced failure node or non-stress-related fault node from the stress function judgment result; and extracts the component identifier of each electrical conduction-affected node and the corresponding cascading failure node or non-cascading failure node from the electrical interlocking function judgment result.

[0142] Edge nodes organize the above three types of judgment results according to component identifiers to form abnormal impact results. The abnormal impact results record the component identifier, affected type, and specific judgment category for each abnormally affected component in list form. The affected types include three types: physical damage, indirect stress wave, and electrical conduction wave. The specific judgment categories include six types: functional failure node, functionally limited node, stress-induced failure node, non-stress-related failure node, cascading failure node, and non-cascading failure node.

[0143] The edge node reads the test result state list corresponding to each component affected by the anomaly from the state mapping relationship, and appends the test result state list to the corresponding component identifier record in the anomaly impact result.

[0144] The embodiments of the present invention enable the abnormal impact results to have an accurate ability to classify the causes of abnormalities by combining the judgment results. This allows the subsequent recycling price assessment to adopt differentiated price deduction strategies for abnormalities with different causes, thereby improving the detection accuracy and price assessment accuracy of the communication terminal detection process.

[0145] Optionally, refer to Figure 2 , Figure 2 This is a schematic diagram of the structure of a communication terminal full-dimensional detection system based on edge-cloud collaboration provided by the present invention. The communication terminal full-dimensional detection system includes: an image recognition module 210, used to control a preceding robotic arm to move the current communication terminal to a preceding detection station, and to perform image recognition on the terminal surface image returned by the vision sensor of the preceding detection station to obtain the image recognition result; a full-dimensional functional testing module 220, used to plan the subsequent robotic arm of the subsequent detection station to perform full-dimensional functional testing based on the image recognition result and the standard detection strategy benchmark corresponding to the model of the current communication terminal, and to obtain functional test data; a full-dimensional state mapping module 230, used to perform state mapping based on the image recognition result and functional test data combined with the three-dimensional digital model of the current communication terminal, and to obtain a full-dimensional state mapping result; and a recycling evaluation module 240, used to match the full-dimensional state mapping result with a recycling price database to obtain a recycling evaluation result for the current communication terminal.

[0146] The embodiments of the present invention improve the detection accuracy of the communication terminal detection process and the accuracy of price evaluation.

[0147] Please see Figure 3 , Figure 3 An embodiment diagram of an electronic device provided in accordance with the present invention. For example... Figure 3 As shown, an embodiment of the present invention provides an electronic device 300, including a memory 310, a processor 320, and a computer program 311 stored in the memory 310 and executable on the processor 320. When the processor 320 executes the computer program 311, it implements the processes of steps 10 to 40.

[0148] Please see Figure 4 , Figure 4 An embodiment diagram of a computer-readable storage medium provided in accordance with an embodiment of the present invention is shown. Figure 4 As shown, this embodiment provides a computer-readable storage medium 400 on which a computer program 311 is stored. When the computer program 311 is executed by a processor, it implements the processes of steps 10 to 40.

[0149] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium, and which, when executed by a processor, implements the processes of steps 10 to 40.

Claims

1. A method for full-dimensional detection of communication terminals based on edge-cloud collaboration, characterized in that, This invention relates to a comprehensive inspection system for communication terminals integrating a robotic arm, edge nodes, and a server cluster. The robotic arm performs the physical transfer of the communication terminals, the edge nodes perform production line scheduling, and the server cluster provides a 3D digital model, standard inspection strategy benchmarks, and a recycling price database. The method includes: The preceding robotic arm is controlled to move the current communication terminal to the preceding inspection station. The visual sensor of the preceding inspection station returns the terminal surface image, and image recognition is performed to obtain the image recognition result. The image recognition result indicates whether there is a defect area on the terminal surface. Based on the image recognition results and the standard detection strategy benchmark corresponding to the current communication terminal model, the subsequent robotic arm of the subsequent detection station performs full-dimensional functional testing to obtain functional test data. Based on the image recognition results and the functional test data, combined with the current three-dimensional digital model of the communication terminal, a state mapping is performed to obtain a full-dimensional state mapping result. The recycling assessment result of the current communication terminal is obtained by matching the full-dimensional state mapping result with the recycling price database.

2. The method for full-dimensional detection of communication terminals based on edge-cloud collaboration according to claim 1, characterized in that, The steps to obtain functional test data include: If the image recognition result indicates that there is no defective area, then a full detection execution instruction is generated based on the standard physical detection trajectory and standard test parameter configuration in the standard detection strategy benchmark; the standard physical detection trajectory covers all preset touch points on the screen, and the standard test parameter configuration includes the standard test process for each functional module; Based on the full-scale detection execution command, the subsequent robotic arm of the subsequent detection station performs full-dimensional functional testing on the current communication terminal to obtain the functional test data.

3. The method for full-dimensional detection of communication terminals based on edge-cloud collaboration according to claim 2, characterized in that, The state mapping based on the image recognition results and the functional test data, combined with the current three-dimensional digital model of the communication terminal, yields a full-dimensional state mapping result, including: If the image recognition result indicates that there is no defective area, then the performance test indicators of each detection point in the functional test data are mapped to the corresponding internal component nodes in the three-dimensional digital model. Based on the spatial adjacency relationship between internal component nodes in the three-dimensional digital model, the spatial dispersion of performance detection indicators of adjacent detection points is analyzed to obtain the performance degradation region. Based on the number of performance degradation regions and the impact of the location of each performance degradation region on the performance of the communication terminal, a full-dimensional state mapping result is generated.

4. The method for full-dimensional detection of communication terminals based on edge-cloud collaboration according to claim 1, characterized in that, The steps to obtain functional test data include: If the image recognition result indicates the existence of a defective area, then the corresponding position points of the defective area in the terminal coordinate system are determined as spatial defect position points. For spatial traversal-type inspection projects, the spatial defect location point is compared with the standard physical inspection trajectory in the standard inspection strategy benchmark. Non-functional related points that overlap with the defect area in the standard physical inspection trajectory are eliminated to obtain the target spatial location point. The detection frequency and detection point density of the associated functional related points adjacent to the defect area in the target spatial location point are adjusted, while keeping the standard functional related points unchanged, to generate a differentiated physical inspection trajectory. The standard functional related points are the points in the target spatial location point other than the associated functional related points. For module-independent testing projects, the corresponding target test parameter configuration is matched in the standard test parameter configuration in the standard testing strategy benchmark based on the spatial proximity between the defect area and each functional module node. Based on the differentiated physical detection trajectory and the target test parameter configuration, an optimized detection execution instruction is generated, and based on the optimized detection execution instruction, the subsequent robotic arm of the subsequent detection station is controlled to perform a full-dimensional functional test on the current communication terminal to obtain the functional test data.

5. The method for full-dimensional detection of communication terminals based on edge-cloud collaboration according to claim 4, characterized in that, The state mapping based on the image recognition results and the functional test data, combined with the current three-dimensional digital model of the communication terminal, yields a full-dimensional state mapping result, including: If the image recognition result indicates the presence of a defective area, the defective area is mapped to the surface mesh of the three-dimensional digital model, and the functional test data is associated with the internal component nodes of the three-dimensional digital model; Based on the spatial topological relationship between the defect region and the internal component nodes in the three-dimensional digital model, the abnormal impact of the defect region on the internal components is determined. Based on the results of the anomaly and the functional test data, a full-dimensional state mapping result is generated.

6. The method for full-dimensional detection of communication terminals based on edge-cloud collaboration according to claim 5, characterized in that, The determination of the abnormal impact of the defect region on the internal components based on the spatial topological relationship between the defect region and the internal component nodes in the three-dimensional digital model includes: Based on the coordinates of each spatial defect location point in the coordinate system of the three-dimensional digital model, a minimum axis-aligned bounding box is constructed to obtain the defect region bounding box. Intersection detection is performed based on the geometric properties of the bounding box of the defect region and each internal component node in the three-dimensional digital model to obtain the intersection detection result; If the intersection detection result indicates that the bounding box of the internal component node has volume overlap or surface contact with the bounding box of the defect area, then the internal component node is identified as a physically damaged node; if the intersection detection result indicates that the bounding box of the internal component node does not overlap with the bounding box of the defect area, then the distance between the center coordinates of the internal component node and the nearest boundary of the bounding box of the defect area is calculated, and internal component nodes whose distance is less than the pre-stored stress transmission distance threshold are identified as indirect stress-affected nodes. The abnormal impact results are obtained by traversing the electrical connection relationships of the internal component nodes in the three-dimensional digital model based on the physically damaged nodes, the indirect stress wave nodes, and the internal component nodes.

7. The method for full-dimensional detection of communication terminals based on edge-cloud collaboration according to claim 6, characterized in that, The abnormal impact results are obtained by traversing the electrical connection relationships of the internal component nodes in the three-dimensional digital model based on physically damaged nodes, indirect stress wave nodes, and the physical damage nodes, including: Based on the electrical connection relationship, traversal is performed with each physically damaged node as the starting node. First-level associated nodes that are directly electrically connected to the starting node are retrieved, and second-level associated nodes that are directly electrically connected to the first-level associated nodes are retrieved until the preset maximum number of electrical cascade layers is reached to obtain the electrical conduction wave nodes. Based on the functional module corresponding to each detection point in the functional test data, the functional module associated with each internal component node in the three-dimensional digital model is matched to obtain the target component node, and the component functional association relationship is obtained based on each detection point and its corresponding target component node. Based on the functional association of the components, the test result status of each detection point is assigned to its corresponding target component node to obtain the status mapping relationship; Based on electrical connection relationships, state mapping relationships, and physical damaged nodes, indirect stress-affected nodes, and electrical conduction-affected nodes, the abnormal impact results are determined.

8. The method for full-dimensional detection of communication terminals based on edge-cloud collaboration according to claim 7, characterized in that, The determination of the abnormal impact results based on electrical connection relationships, state mapping relationships, combined with physically damaged nodes, indirect stress-affected nodes, and electrically conducted-affected nodes includes: Based on the state mapping relationship, state matching is performed. If the test result of the physically damaged node is all failures, the physically damaged node is determined as a functional failure node; otherwise, the physically damaged node is determined as a functionally limited node, and the damaged function determination result is obtained. Based on the state mapping relationship, state matching is performed. If the test result of the indirect stress-affected node is a failure and the electrical connection relationship indicates that there is at least one directly electrical connection physically damaged node, then the indirect stress-affected node is determined as a stress-induced failure node. Otherwise, the indirect stress-affected node is determined as a non-stress-related fault node, and the stress function determination result is obtained. Based on the state mapping relationship, state matching is performed. If the test result of the electrical conduction wave node is a failure, and the electrical connection relationship indicates that the upstream dependent node of the electrical conduction wave node is a physically damaged node or the upstream dependent node includes at least one functionally failed node, then the electrical conduction wave node is determined as a cascading failure node; otherwise, the electrical conduction wave node is determined as a non-cascading failure node, and the electrical cascading function determination result is obtained. The abnormal impact result is obtained by combining the damage function determination result, the stress function determination result, and the electrical interlock function determination result.

9. The method for full-dimensional detection of communication terminals based on edge-cloud collaboration according to any one of claims 1 to 8, characterized in that, The method further includes: Monitor the state of the first region of the first buffer corresponding to the output of the preceding detection station, and the state of the second region of the second buffer corresponding to the input of the following detection station; If the preceding detection station where the current communication terminal is located completes the test, and the second area of ​​the second buffer is in an idle state, a release command is sent to the robotic arm used to transfer the current communication terminal, so as to control the robotic arm to transfer the current communication terminal to the second buffer through the release command; If there is a communication terminal to be processed in the first buffer and the first area of ​​the preceding detection station is in an idle state, a gripping command is sent to the robotic arm used for loading, so as to control the robotic arm to move the next communication terminal from the first buffer to the preceding detection station through the gripping command.

10. A multi-dimensional detection system for communication terminals based on edge-cloud collaboration, characterized in that, The system is used to implement the end-edge-cloud collaborative full-dimensional detection method for communication terminals as described in any one of claims 1 to 9; the system includes: The image recognition module is used to control the preceding robotic arm to move the current communication terminal to the preceding inspection station, perform image recognition on the terminal surface image returned by the vision sensor of the preceding inspection station, and obtain the image recognition result; the image recognition result indicates whether there is a defect area on the terminal surface. The full-dimensional functional testing module is used to plan the subsequent robotic arm of the subsequent testing station based on the image recognition results and the standard detection strategy benchmark corresponding to the current communication terminal model to perform full-dimensional functional testing and obtain functional test data. The full-dimensional state mapping module is used to perform state mapping based on the image recognition results and the functional test data, combined with the current three-dimensional digital model of the communication terminal, to obtain the full-dimensional state mapping result. The recycling assessment module is used to match the full-dimensional state mapping results with the recycling price database to obtain the recycling assessment result of the current communication terminal.

Images