Material anti-theft method based on material type fingerprint texture pattern

Abstract

The present disclosure provides a material anti-theft method based on material type fingerprint texture pattern, comprising: scanning vehicle-mounted materials at a starting position and a destination position of material transportation by sensors respectively to obtain material three-dimensional form information at the corresponding positions; preprocessing the material three-dimensional form information to obtain contour line texture maps; analyzing and comparing the contour line texture maps at the starting position and the destination position, and performing similarity calculation; executing a preset similarity logic on the similarity calculation result, or analyzing and comparing the similarity calculation result with a similarity threshold to determine whether the material form has undergone abnormal form change or the material has been stolen during transportation. The present disclosure can effectively monitor the change degree of the material during transportation, timely discover the case of material theft or tampering, and improve the safety and reliability of material transportation.

Description

Technical Field

[0001] This disclosure relates to the field of material transportation safety monitoring technology, specifically to a material anti-theft method based on material shape fingerprint texture pattern. Background Technology

[0002] With the rapid development of the logistics and transportation industry, security issues during material transportation have become increasingly prominent. Especially in the transportation of bulk materials such as grain and coal, theft or substitution of materials occurs frequently, causing significant economic losses to enterprises. Currently, theft prevention and monitoring during material transportation mainly rely on traditional methods such as manual inspection and weighing comparison. These methods suffer from low efficiency, large errors, and susceptibility to human factors.

[0003] Furthermore, in practical applications, theft prevention monitoring during material transportation faces the following challenges: First, materials such as grain or coal are frequently stolen after being loaded at the starting point and transported to the destination, sometimes even resulting in the replacement of entire truckloads of materials. Second, it is difficult to detect theft when materials are identical but have similar shapes. Third, the replacement of entire truckloads of materials is even more difficult to detect. Fourth, most surface deformation detection currently relies on human visual inspection, which is highly subjective, and human eyes are prone to fatigue, leading to high rates of false positives and false negatives. Fifth, large material yards have large material transportation volumes and numerous vehicles, making manual inspection prone to loopholes or errors, and making it difficult to detect theft in a timely manner.

[0004] Therefore, there is an urgent need in this field for a technical solution that can solve the above-mentioned technical problems. Summary of the Invention

[0005] This disclosure provides a material anti-theft method based on material fingerprint texture patterns, which is used to solve the technical problems of existing vehicles transporting materials from the starting position to the destination, such as untimely monitoring of whether the materials have been stolen or the inability to effectively determine whether the materials transported by the vehicle have been stolen.

[0006] This disclosure provides a material anti-theft method based on material fingerprint texture patterns, including the following steps:

[0007] Sensors scan the onboard materials at the starting and ending points of material transportation to obtain the three-dimensional morphological information of the materials at the corresponding locations.

[0008] The contour texture map is obtained by preprocessing the three-dimensional morphology information of the material.

[0009] Analyze and compare the contour texture maps of the starting position and the arriving position, and calculate the similarity.

[0010] The similarity calculation results are executed with preset similarity logic, or the similarity calculation results are compared with similarity threshold analysis to determine whether the material form has undergone abnormal changes during transportation or whether the material has been stolen.

[0011] According to at least one embodiment of the material anti-theft method based on material fingerprint texture pattern, the analysis and comparison of contour texture maps of the starting position and the arriving position, and the similarity calculation, include the following steps:

[0012] Analyze the contour texture map at the starting position to obtain the centroid or center position information of at least one set of concentric closed curves in the contour texture map;

[0013] Analyze the contour texture map at the arrival location to obtain the centroid or center position information of at least one set of concentric closed curves in the contour texture map;

[0014] Compare the centroid or center position information change vectors of at least one set of concentric closed curves at the starting position and the reached position;

[0015] Similarity is calculated based on the change vector of the centroid or center position information of at least one set of concentric closed curves.

[0016] According to at least one embodiment of the material anti-theft method based on material fingerprint texture pattern, the analysis and comparison of contour texture maps of the starting position and the arriving position, and the similarity calculation, include the following steps:

[0017] Analyze the contour texture map at the starting position to obtain the centroid or center position information of each concentric closed curve in each group of concentric closed curves in the contour texture map;

[0018] Analyze the contour texture map at the arrival location to obtain the centroid or center position information of each concentric closed curve in each group of concentric closed curves in the contour texture map;

[0019] Compare the centroid or center position change vector of each concentric closed curve in each set of concentric closed curves at the starting position and the reached position;

[0020] The similarity is calculated based on the change vector of the centroid or center position of each concentric closed curve in each group of concentric closed curves.

[0021] According to at least one embodiment of the material anti-theft method based on material fingerprint texture pattern, the analysis and comparison of contour texture maps of the starting position and the arriving position, and the similarity calculation, include the following steps:

[0022] Analyze the contour texture map at the starting position to obtain the centroid or center position information of each set of concentric closed curves in the contour texture map.

[0023] Analyze the contour texture map at the arrival location to obtain the centroid or center position information of each set of concentric closed curves in the contour texture map.

[0024] Compare the centroid or center position change vector of each set of concentric closed curves in the set of concentric closed curves at the starting position and the set of reached position;

[0025] Similarity is calculated based on the change vector of the centroid or center position of each of the multiple concentric closed curves.

[0026] According to at least one embodiment of the material anti-theft method based on material fingerprint texture pattern, the analysis and comparison of contour texture maps of the starting position and the arriving position, and the similarity calculation, include the following steps:

[0027] Obtain the centroid or center position information of multiple sets of concentric closed curves in the contour texture map corresponding to the starting position and the arriving position;

[0028] Calculate the relative position change vector of the centroid or center of each pair of concentric closed curves from the starting position to the reached position;

[0029] Similarity is calculated based on the relative position change vector of the centroid or center of each pair of concentric closed curves.

[0030] According to at least one embodiment of the material anti-theft method based on material fingerprint texture pattern of this disclosure, the method further includes the following steps in analyzing and comparing the contour texture maps of the starting position and the arriving position and performing similarity calculation:

[0031] Obtain multiple sets of concentric closed curves in the contour texture map corresponding to the starting position and the arriving position;

[0032] Analyze and compare the spacing change vector between two adjacent concentric closed curves in each group of concentric closed curves, and calculate the similarity based on the spacing change vector.

[0033] According to at least one embodiment of the material anti-theft method based on material fingerprint texture pattern, before scanning the vehicle-mounted material at the starting and arrival positions of material transportation using sensors to obtain the three-dimensional morphological information of the material at the corresponding positions, the method further includes the following steps:

[0034] At the starting position and the arrival position, the vehicle transporting the material is scanned by a camera device to obtain the first vehicle feature information.

[0035] According to at least one embodiment of the present disclosure, the material anti-theft method based on material fingerprint texture pattern scans the vehicle-mounted material at the starting position and the arrival position of the material transportation by means of sensors to obtain the three-dimensional shape information of the material at the corresponding position, and then further includes the following steps.

[0036] The material's three-dimensional morphology information is grouped and set according to the first vehicle feature information;

[0037] Obtain at least one set of three-dimensional morphological information of the material corresponding to the first vehicle feature information.

[0038] According to at least one embodiment of the present disclosure, a material anti-theft method based on material fingerprint texture patterns further includes the following steps:

[0039] Obtain the material mass information and / or material volume information from the three-dimensional morphology information of the material;

[0040] Analyze and compare the material quality information and / or material volume information at the starting position and the arriving position to determine whether the material was stolen during transportation.

[0041] The material anti-theft method based on material fingerprint texture pattern according to at least one embodiment of the present disclosure further includes the following steps:

[0042] If it is determined that the material has changed during transportation, or if it is determined that the degree of change in the material during transportation exceeds a set threshold, a warning will be issued that the material has been stolen.

[0043] The material anti-theft method based on material fingerprint texture pattern according to at least one embodiment of the present disclosure further includes the following steps:

[0044] The sensors scan the material transport vehicle at the starting position and the arrival position respectively to obtain the vehicle's second vehicle characteristic information;

[0045] Based on the three-dimensional morphological information of the material and the second vehicle feature information, a contour texture map of the integrated vehicle and material is obtained through preprocessing.

[0046] According to at least one embodiment of the material anti-theft method based on material fingerprint texture pattern, the method for obtaining a contour texture map of the vehicle and material integrated by preprocessing the three-dimensional morphological information of the material and the second vehicle feature information includes the following steps:

[0047] The material area range is defined by the vehicle frame boundary information containing the material in the second vehicle feature information.

[0048] A reference coordinate system is established based on the vehicle frame boundary information, and the point cloud data information in the three-dimensional shape information of the material corresponding to the starting position and the arrival position of the material transportation is converted to the reference coordinate system.

[0049] The point cloud data information in the converted three-dimensional morphological information of the material is combined with the material area range for preprocessing to obtain a contour texture map of the integrated vehicle and material.

[0050] The material anti-theft method based on material fingerprint texture pattern disclosed herein has the following advantages compared with the prior art:

[0051] 1. This disclosure establishes contour maps of the material at the starting and ending positions, compares the contour maps after loading and before unloading, calculates the similarity, and analyzes and compares them with a similarity threshold to determine whether the material has changed during transportation and the extent of the change.

[0052] 2. This disclosure can scan the vehicle transporting materials using a camera device to obtain the first vehicle feature information, which mainly includes the vehicle license plate, vehicle color, vehicle outline, and outline of the materials being transported. Then, after determining that the materials have changed or the degree of change during transportation, the first vehicle feature information can be combined to provide a warning prompt for checking the vehicle and materials, thereby greatly saving the time of manual inspection of vehicles and materials.

[0053] 3. This disclosure can also use sensors to scan vehicle information at the starting and ending positions to obtain second vehicle feature information. The second vehicle feature information and the three-dimensional morphology information of the material are preprocessed to obtain a contour texture map of the vehicle and material as a whole. Thus, vehicle and material information can be analyzed at the same time to determine whether the vehicle has been replaced or whether the material has been stolen.

[0054] 4. This disclosure converts the point cloud data information in the three-dimensional shape information of the material to the reference coordinate system established by the vehicle boundary information, thereby greatly improving the accuracy of material shape recognition after introducing vehicle feature information.

[0055] 5. After establishing the contour texture map, this disclosure can convert the three-dimensional morphological information of the material into a material fingerprint map. Combined with the material mass and volume information in the three-dimensional morphological information of the material, it can quickly and efficiently determine whether the material has been stolen, and can efficiently monitor and identify transport vehicles and materials.

[0056] 6. This disclosure achieves accurate judgment on whether materials have been stolen during transportation by calculating the similarity of the centroid or center position change vector of concentric closed curves in the contour texture map, or the distance change vector between adjacent concentric closed curves. Attached Figure Description

[0057] The accompanying drawings are provided to further illustrate the present disclosure and form part of the specification. They are used together with the following detailed description to explain the present disclosure, but do not constitute a limitation thereof. In the drawings:

[0058] Figure 1 This is a schematic flowchart of a material anti-theft method based on material shape fingerprint texture pattern according to an embodiment of this disclosure. Figure 1 ;

[0059] Figure 2 This is a schematic flowchart of a material anti-theft method based on material shape fingerprint texture pattern according to an embodiment of this disclosure. Figure 2 ;

[0060] Figure 3 This is a schematic diagram showing the status of the vehicles transporting materials and the materials themselves.

[0061] Figure 4 yes Figure 3 A schematic diagram of the contour texture of the material in the middle;

[0062] Figure 5 This is a comparative schematic diagram of the material contour texture maps at the starting and ending positions;

[0063] Figures 6-8 It is a schematic diagram of the contour lines of the material at the starting and destination positions during actual transportation. Detailed Implementation

[0064] The specific embodiments of this disclosure will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit this disclosure.

[0065] like Figure 1 As shown, this disclosure provides a material anti-theft method based on material fingerprint texture patterns, including the following steps:

[0066] S200: At the starting and ending points of material transportation, sensors scan the on-board materials to obtain the three-dimensional morphological information of the materials at the corresponding locations.

[0067] S400: Preprocess the material's three-dimensional morphology information to obtain a contour texture map;

[0068] S600. Analyze and compare the contour texture maps of the starting position and the arriving position, and calculate the similarity.

[0069] S800: Execute preset similarity logic on the similarity calculation result, or compare the similarity calculation result with the similarity threshold analysis to determine whether the material form has undergone abnormal changes during transportation, or to determine whether the material has been stolen.

[0070] This disclosure provides a material anti-theft method based on material shape fingerprint texture pattern. By collecting the three-dimensional shape information of the material at the starting position and the arrival position of the material transportation, and by comparing and analyzing this information, it can be determined whether the shape of the material has undergone abnormal changes or whether the material has changed during transportation, thereby achieving the purpose of material anti-theft.

[0071] Next, sensors scan the onboard materials at the starting and ending points of the material transport to obtain the three-dimensional morphological information of the materials at the corresponding locations. These sensors can be lidar, millimeter-wave radar, or other devices capable of acquiring three-dimensional morphological information. The three-dimensional morphological information of the materials includes their shape, volume, surface contour, and other three-dimensional spatial information.

[0072] In this disclosure, the three-dimensional morphological information of the material is preprocessed to obtain a contour map. The contour map is a two-dimensional representation of the material's surface morphology, consisting of a set of closed curves formed by connecting points of equal height on the material's surface. The contour map can visually reflect the surface morphological characteristics of the material, similar to contour lines on a topographic map.

[0073] In a further embodiment of this disclosure, the sensor may employ a high-precision lidar with a scanning accuracy down to the millimeter level, capable of accurately capturing minute changes on the material surface. The three-dimensional morphological information of the material includes point cloud data of the material surface, with each point containing three-dimensional coordinate information (x, y, z).

[0074] During the preprocessing process, the point cloud data is first filtered to remove noise points and outliers; then, point cloud downsampling is performed to reduce the amount of data and improve processing efficiency; next, point cloud registration is performed to align point cloud data collected from different locations to the same coordinate system; finally, a contour texture map is generated based on the point cloud data, and the contour interval can be set to 5 cm to clearly display the morphological features of the material surface.

[0075] Similarity can be calculated using the Structural Similarity Index (SSIM) method, which comprehensively considers similarity in three aspects: brightness, contrast, and structure. The calculation formula is as follows:

[0076] SSIM(x,y)=[l(x,y)] α ·[c(x,y)]β ·[s(x,y)] γ ;

[0077] Where l, c, and s represent brightness, contrast, and structural similarity, respectively, and α, β, and γ are weighting coefficients, all set to 1. Specifically, the similarity threshold can be set to 0.85, meaning that when the calculated similarity is below 0.85, it is determined that the material has changed during transportation. When the similarity is between 0.85 and 0.95, it is judged as a slight change; when the similarity is below 0.85, it is judged as a significant change, and there may be evidence of material theft.

[0078] Alternatively, similarity calculation can employ a multi-feature fusion method, comprehensively considering the shape, topological, and statistical features of contour lines. Shape features include the area, perimeter, and circularity of contour lines; topological features include the nesting relationships and connectivity of contour lines; and statistical features include the height and density distribution of contour lines. These features form a high-dimensional feature vector, and similarity is measured by calculating the Euclidean distance between these feature vectors. Specifically, the similarity threshold can be set to a distance value of 30 and dynamically adjusted according to different types of materials. For regularly shaped materials, such as sand and coal, a lower threshold is set; for irregularly shaped materials, such as scrap metal and wood, a higher threshold is set.

[0079] Specifically, such as Figure 2 As shown, before scanning the onboard materials at the starting and ending points of material transportation using sensors to obtain the three-dimensional morphological information of the materials at the corresponding locations, the following steps are also included:

[0080] S100: At the starting position and the arrival position, the vehicle transporting the material is scanned by a camera device to obtain the first vehicle feature information of the vehicle.

[0081] The camera device can employ a high-definition camera array, including multiple cameras at different angles, to capture omnidirectional images of the vehicle. The first vehicle feature information includes the license plate number, vehicle color, model, dimensions, and special markings. The license plate number is automatically identified using Optical Character Recognition (OCR) technology, while the vehicle color and model are automatically classified using a deep learning model. This first vehicle feature information can be used for subsequent grouping and matching of material 3D morphology information.

[0082] Alternatively, the camera system can be an intelligent traffic camera system, including license plate recognition cameras and panoramic cameras. License plate recognition cameras are specifically designed to capture license plate images and identify license plate numbers, while panoramic cameras capture the overall appearance of the vehicle. In addition to basic information such as the license plate number and vehicle model, the primary vehicle characteristic information also includes the vehicle's unique identifier (such as RFID tag information), transportation company information, and driver information.

[0083] After scanning the onboard materials at the starting and ending points of material transportation using sensors to obtain the corresponding three-dimensional morphological information of the materials, the process further includes the following steps: S300, grouping the three-dimensional morphological information of the materials according to the first vehicle feature information; obtaining at least one group of the three-dimensional morphological information of the materials corresponding to the first vehicle feature information, such as... Figure 3 As shown.

[0084] In this disclosure, after obtaining the three-dimensional morphological information of the material, the three-dimensional morphological information of the material is grouped according to the first vehicle feature information, and at least one set of material three-dimensional morphological information corresponding to the first vehicle feature information is obtained. This ensures that the comparative analysis is of material information of the same vehicle at different locations, avoiding confusion.

[0085] In the process of grouping the three-dimensional morphological information of materials based on the first vehicle feature information, a vehicle information database is first established to record the feature information of each vehicle; then, the three-dimensional morphological information of materials obtained from each scan is associated with the corresponding vehicle feature information to form a "vehicle-material" pairing; finally, the three-dimensional morphological information of materials is grouped according to the unique identifier of the vehicle (such as the license plate number) to ensure that the comparison is of material information of the same vehicle in different locations.

[0086] When acquiring at least one set of material 3D morphology information corresponding to the first vehicle's feature information, the system queries the database to find historical scan records that match the current vehicle. If it is the first scan of the vehicle, its information is used as the baseline record; if it is a vehicle with existing records, the current scan result is compared with the most recent scan result.

[0087] In addition, the system will determine whether the current scan is a start-point scan or an arrival-point scan based on the vehicle's transportation route and time information. For vehicles that regularly travel on fixed routes, the system will automatically identify their transportation mode and adjust the comparison strategy accordingly.

[0088] In a further embodiment of this disclosure, the above-mentioned material anti-theft method based on material fingerprint texture pattern further includes the following steps: obtaining material mass information and / or material volume information from the three-dimensional morphology information of the material; analyzing and comparing the material mass information and / or material volume information at the starting position and the arriving position to determine whether the material has been stolen during transportation.

[0089] In other words, in addition to similarity calculation through contour texture maps, the method disclosed herein can also obtain material quality information and / or material volume information from the three-dimensional morphological information of the material, analyze and compare the material quality information and / or material volume information at the starting position and the destination position, so as to determine whether the material has been stolen during transportation.

[0090] Alternatively, the material's mass information can also be obtained through a weight sensor, i.e., this disclosure uses both a sensor and a weight sensor structure. Specifically, the weight sensor is installed under the vehicle chassis and can measure the total weight of the materials on board. The material volume information is obtained by calculating the volume of three-dimensional point cloud data using the three-dimensional convex hull volume calculation method. In comparing the material mass information, a mass change threshold of 3% of the total material mass is set. If the mass of the material measured at the arrival location is more than 3% less than the starting location, it is determined that the material may have been stolen. Simultaneously, in comparing the material volume information, a volume change threshold of 5% of the total material volume is set. If the volume of the material measured at the arrival location is more than 5% less than the starting location, it is also determined that the material may have been stolen. Therefore, the comparison results of mass and volume information can be combined with the contour map similarity calculation results for a comprehensive judgment. If the mass or volume reduction exceeds the threshold, and the contour map similarity is lower than the similarity threshold, theft is highly suspected, and the system will issue a high-level alarm; if only one indicator exceeds the threshold, a medium-level alarm is issued; if both indicators are within the threshold range, the material is considered safe.

[0091] This disclosure primarily uses the similarity calculation of contour texture maps, and can also combine material quality information and / or material volume information at the starting and ending positions, thereby enabling a more effective and accurate determination of whether materials have been stolen during transportation.

[0092] like Figure 5-8 As shown, this disclosure mainly involves analyzing and comparing the contour texture maps of the starting and ending positions, and then calculating the similarity. Several specific embodiments of the similarity calculation method are described below.

[0093] 1) Analyzing and comparing the contour texture maps of the starting position and the arriving position, and calculating the similarity, includes the following steps: analyzing the contour texture map at the starting position to obtain the centroid or center position information of at least one set of concentric closed curves in the contour texture map; analyzing the contour texture map at the arriving position to obtain the centroid or center position information of at least one set of concentric closed curves in the contour texture map; comparing the change vector of the centroid or center position information of at least one set of concentric closed curves at the starting position and the arriving position; and calculating the similarity based on the change vector of the centroid or center position information of at least one set of concentric closed curves.

[0094] In this embodiment, the contour texture maps of the starting position and the destination position are analyzed to obtain the centroid or center position information of at least one set of concentric closed curves in the contour texture map, the change vectors of the centroid or center position information of the concentric closed curves at the two positions are compared, and similarity is calculated based on these change vectors.

[0095] like Figure 4 and 5 As shown, the materials transported by the vehicle at the origin and destination positions contain four convex hulls. It can be assumed that the materials form five sets of concentric closed curves, with their centers or centroids located at A, B, C, D, E and A', B', C', D', E' respectively before and after transport. Figure 5 As can be seen, all five sets of concentric closed curves shifted in the same direction and proportionally. Figure 5 The letters A, B, C, D and A', B', C', D' correspond to Figure 4 The four convex hulls, E and E', correspond to the contour lines of the material at the bottom of the vehicle.

[0096] In this embodiment, the sensor can be a millimeter-wave radar with an operating frequency of 77 GHz, and its scanning range can cover the entire material area on the vehicle. In addition to point cloud data, the three-dimensional morphology information of the material also includes reflection intensity information, which can better distinguish materials of different textures.

[0097] In the process of generating contour texture maps, point cloud data is first projected onto a horizontal plane to form a height map; then contour lines are generated based on the height map, with the contour line interval set to 10 centimeters; finally, the contour lines are smoothed to reduce jagged edges and improve the quality of the texture map.

[0098] When obtaining the centroid or center position information of concentric closed curves, the centroid calculation method is used. For each closed curve, the average coordinates of all its points are calculated as the centroid position. Specifically, assuming there are n points on the closed curve with coordinates (x1, y1), (x2, y2), ..., (xn, yn), the centroid coordinates are (∑xi / n, ∑yi / n).

[0099] When calculating the change vector of the center of gravity position information, for each pair of concentric closed curves corresponding to the starting position and the destination position, the difference vector of their center of gravity positions is calculated. Assuming that the center of gravity coordinates of a closed curve at the starting position are (x1, y1) and the center of gravity coordinates of the closed curve at the destination position are (x2, y2), then the change vector is (x2-x1, y2-y1).

[0100] The similarity calculation uses the cosine similarity method, and the calculation formula is: cos(θ)=(A·B) / (|A|·|B|), where A and B are the feature vectors of the starting position and the destination position, respectively, and are composed of the coordinates of the centroids of each closed curve.

[0101] One approach is to set the similarity threshold to 0.9. When the calculated similarity is below 0.9, it is determined that the material has changed during transportation.

[0102] 2) Analyzing and comparing the contour texture maps of the starting position and the arriving position, and calculating the similarity, includes the following steps: analyzing the contour texture map at the starting position, obtaining the centroid or center position information of each concentric closed curve in each group of concentric closed curves in the contour texture map; analyzing the contour texture map at the arriving position, obtaining the centroid or center position information of each concentric closed curve in each group of concentric closed curves in the contour texture map; comparing the centroid or center position change vector of each concentric closed curve in each group of concentric closed curves at the starting position and the arriving position; and calculating the similarity based on the centroid or center position change vector of each concentric closed curve in each group of concentric closed curves.

[0103] In this embodiment, the contour texture maps of the starting position and the destination position are analyzed to obtain the centroid or center position information of each concentric closed curve in each group of concentric closed curves in the contour texture map. The centroid or center position change vector of each concentric closed curve in each group of concentric closed curves at the two positions is compared, and similarity is calculated based on these change vectors.

[0104] In this embodiment, the sensor can be a 3D LiDAR array, composed of multiple LiDARs, which can scan the material from different angles, reducing occlusion issues and improving the integrity of the scan. The scanning frequency is 10Hz, and a single scan can acquire point cloud data of approximately 1 million points.

[0105] The contour texture map is generated using an adaptive contour spacing method, which automatically adjusts the contour spacing according to the complexity of the material surface. For areas with drastic surface changes, the contour spacing is smaller, allowing for a more detailed description of surface features; for flat areas, the contour spacing is larger, reducing redundant information.

[0106] In the process of obtaining concentric closed curves, local extreme points (peaks or valleys) in the contour texture map are first identified. Using these points as centers, groups of concentric closed curves are extracted. For each group of concentric closed curves, they are numbered sequentially from the inside out, and the centroid position of each closed curve is calculated.

[0107] When calculating the vector of change in the center of gravity position, for each set of concentric closed curves corresponding to the starting position and the ending position, the change in the center of gravity position of the closed curves with the same number is compared. This allows for a more precise capture of changes in the surface morphology of the material.

[0108] Similarity calculation uses a weighted Euclidean distance method, assigning different weights to closed curves at different heights. Generally, closed curves at the top of the material have a larger weight because the top is more easily altered; closed curves at the bottom of the material have a smaller weight because the bottom is usually more stable.

[0109] The similarity threshold can be set to a distance value of 20. When the calculated weighted Euclidean distance is greater than 20, it is determined that the material has undergone significant changes during transportation.

[0110] 3) Analyzing and comparing the contour texture maps of the starting position and the arriving position, and performing similarity calculation, includes the following steps: analyzing the contour texture map at the starting position, obtaining the centroid or center position information of each of the multiple sets of concentric closed curves in the contour texture map; analyzing the contour texture map at the arriving position, obtaining the centroid or center position information of each of the multiple sets of concentric closed curves in the contour texture map; comparing the centroid or center position change vector of each of the multiple sets of concentric closed curves at the starting position and the arriving position; and performing similarity calculation based on the centroid or center position change vector of each of the multiple sets of concentric closed curves.

[0111] In this embodiment, the contour texture map of the starting position and the destination position is analyzed to obtain the centroid or center position information of each of the multiple concentric closed curves in the contour texture map. The centroid or center position change vector of each of the multiple concentric closed curves at the two positions is compared, and similarity is calculated based on these change vectors.

[0112] In this embodiment, the sensor can be a phased array radar with an operating frequency of 24 GHz, enabling electronic scanning without mechanical rotation, resulting in fast scanning speed and high stability. The scanning range covers the entire vehicle-mounted material area, with a scanning resolution of 0.5 degrees.

[0113] In the process of generating contour texture maps, the point cloud data is first segmented to separate the material area from the non-material area; then the point cloud data of the material area is meshed to generate a regular grid; next, the grid is smoothed to reduce the impact of noise; finally, the contour texture map is generated based on the smoothed grid, with the contour line interval set to 8 cm.

[0114] In the identification of multiple sets of concentric closed curves, the density-based clustering algorithm DBSCAN was used to automatically identify different peaks in the contour texture map. Figure 4 Each region consists of four convex hulls or valley regions (the lowest points between the convex hulls), and each region corresponds to a set of concentric closed curves. For each set of concentric closed curves, its geometric center is calculated as the center position information of that set.

[0115] When calculating the vector of change in the center position, the first step is to solve the matching problem of concentric closed curve groups at the starting and ending positions. A feature-descriptor-based matching method is used, where a set of feature descriptors is calculated for each group of concentric closed curves, including area, perimeter, and shape factor, and then matching is performed based on the similarity of the descriptors.

[0116] The similarity calculation uses the Mahalanobis distance method, which considers the correlation between different features. Specifically, the similarity threshold can be set to a distance value of 15. When the calculated Mahalanobis distance value is greater than 15, it is determined that the material has changed during transportation. In addition, a grading standard for the degree of change is set: a distance value between 15 and 25 indicates a slight change, a distance value between 25 and 40 indicates a moderate change, and a distance value greater than 40 indicates a severe change.

[0117] 4) Analyzing and comparing the contour texture maps of the starting position and the destination position, and performing similarity calculation includes the following steps: obtaining the centroid or center position information of multiple sets of concentric closed curves in the contour texture maps corresponding to the starting position and the destination position; calculating the relative position change vector of the centroid or center of each pair of concentric closed curves from the starting position to the destination position; and performing similarity calculation based on the relative position change vector of the centroid or center of each pair of concentric closed curves.

[0118] In this embodiment, the centroid or center position information of multiple sets of concentric closed curves at the starting position and the destination position is obtained, the relative position change vector of the centroid or center of each pair of concentric closed curves from the starting position to the destination position is calculated, and the similarity is calculated based on the relative position change vector of the centroid or center of each pair of concentric closed curves.

[0119] In this embodiment, the sensor can employ a dual-frequency radar system, using both a 77GHz high-frequency radar and a 24GHz low-frequency radar. The high-frequency radar provides high-resolution close-range scanning, while the low-frequency radar provides long-range scanning coverage. The combination of the two can obtain more comprehensive three-dimensional morphological information of the material.

[0120] Contour texture maps can be generated using a multi-resolution approach, applying different resolutions to different areas of the material surface. High resolution is used for areas with rich detail, while low resolution is used for flat areas. This approach improves processing efficiency while maintaining accuracy.

[0121] When obtaining the centroid or center position information of multiple sets of concentric closed curves, an image moment-based method is used. For each set of concentric closed curves, its zeroth moment (area) and first moment (centroid) are calculated as feature information of that set.

[0122] When calculating the relative position change vector of the centroid or center of every two sets of concentric closed curves, first select n significant concentric closed curves in the contour texture map at the starting position, and then find the corresponding n concentric closed curves in the contour texture map at the destination position. For any two sets i and j, calculate their relative position vector of the centroid at the starting position: V ij _start=(x i _start-x j _start,y i _start-y j _start), and the relative position vector V of the centroid at the arrival position. ij _end=(x i _end-x j _end,y i _end-y j Then calculate the difference ΔV between the two vectors. ij =V ij _end-V ij _start serves as the vector representing the relative position change of the center of gravity.

[0123] The similarity calculation uses the vector similarity method, calculating all ΔV. ij The average magnitude of the vectors is used as a similarity index. The similarity threshold can be set to 5 cm. When the calculated average magnitude is greater than 5 cm, it is determined that the material has changed during transportation. In addition, the consistency of the direction of the changing vectors is also considered. If the direction of the changing vectors differs greatly between different groups, it indicates that the material morphology has undergone complex changes, which may be the result of human intervention.

[0124] 5) Analyzing and comparing the contour texture maps of the starting position and the arriving position, and performing similarity calculation, further includes the following steps: obtaining multiple sets of concentric closed curves in the contour texture maps corresponding to the starting position and the arriving position; analyzing and comparing the spacing change vector between two adjacent concentric closed curves in each set of concentric closed curves in the contour texture map, and performing similarity calculation based on the spacing change vector.

[0125] In this embodiment, the spacing change vector between two adjacent concentric closed curves in each group of concentric closed curves in the contour texture map is analyzed and compared, and similarity is calculated based on the spacing change vector.

[0126] In this embodiment, after obtaining at least one set of concentric closed curves in the contour texture map based on the above methods 1)-4), the spacing change vector between each pair of adjacent concentric closed curves at the starting position and the ending position can be calculated respectively. For example, the spacing change vector between two adjacent concentric closed curves at multiple set directions can be calculated, and then similarity calculation can be performed.

[0127] In a further embodiment of this disclosure, the above-mentioned material anti-theft method based on material fingerprint texture pattern further includes the following steps: if it is determined that the material has changed during transportation, or if it is determined that the degree of change of the material during transportation exceeds a set threshold, then a material theft warning is issued.

[0128] This disclosure applies preset similarity logic to the similarity calculation results, or compares the similarity calculation results with a similarity threshold analysis, to determine whether the material's form has undergone abnormal changes during transportation, or whether the material has been stolen. If it is determined that the material has changed during transportation, or that the degree of change exceeds a set threshold, a theft warning is issued. The warning is divided into three levels: slight change (yellow alert), moderate change (orange alert), and severe change (red alert). Different levels of alerts trigger different processing flows, such as logging, notifying the administrator, activating camera recording, notifying security personnel, and notifying manual inspection of vehicles and materials. Furthermore, the system generates a detailed change report, including visual markers of the changed area, estimates of the change amount, and inferences about the change time, helping managers quickly locate the problem and take appropriate measures.

[0129] In addition, the above-mentioned material anti-theft method based on material fingerprint texture patterns may also include the following steps:

[0130] S500: At the starting position and the arrival position, the vehicle transporting the material is scanned by the sensor to obtain the second vehicle feature information of the vehicle; the material three-dimensional morphology information and the second vehicle feature information are preprocessed to obtain a contour texture map of the vehicle and material as a whole.

[0131] In this disclosure, the sensor scans not only the material but also the vehicle transporting the material, acquiring secondary vehicle feature information. This secondary vehicle feature information includes the vehicle's three-dimensional outline, cargo box dimensions, and frame structure, which can be obtained from radar scan data using point cloud segmentation and feature extraction algorithms.

[0132] Preprocessing the material's three-dimensional morphology information and the second vehicle feature information to obtain a contour texture map integrating the vehicle and material includes the following steps: using the vehicle frame boundary information containing the material in the second vehicle feature information as the material area; establishing a reference coordinate system based on the vehicle frame boundary information, and converting the point cloud data information in the material's three-dimensional morphology information corresponding to the starting position and the arrival position of the material transportation to the reference coordinate system; combining the converted point cloud data information in the material's three-dimensional morphology information with the material area range to obtain a contour texture map integrating the vehicle and material, such as... Figure 4-5 As shown.

[0133] The aforementioned integrated contour map of the vehicle and materials contains comprehensive information about both, providing a more accurate reflection of the material distribution on the vehicle. When calculating similarity, the vehicle portion is first registered to ensure that materials in the same location are being compared; then, the similarity of the material portion is calculated to eliminate the influence of vehicle differences.

[0134] Therefore, in the specific embodiments of the five methods described above that analyze and compare the contour texture maps of the starting position and the destination position and perform similarity calculations, the contour map can be a contour texture map of only the material itself, or it can be a contour texture map of the material and the vehicle as a whole.

[0135] Furthermore, the similarity threshold can be dynamically adjusted based on different types of materials and vehicles. For materials transported in open compartments (such as flatbed trucks), the threshold is set lower because changes in material shape are easier to detect; for materials transported in closed compartments, the threshold is set higher to accommodate the limitations of detection conditions.

[0136] When calculating similarity, the characteristics of different types of vehicles and materials must be considered. For example, for vehicles transporting bulk materials (such as coal and sand), the focus should be on changes in the volume and shape of the materials; for vehicles transporting containers, the focus should be on the integrity and sealing of the containers.

[0137] This disclosure can effectively monitor the degree of change of materials during transportation, promptly detect cases of material theft or tampering, and improve the safety and reliability of material transportation.

[0138] The preferred embodiments of this disclosure have been described in detail above with reference to the accompanying drawings. However, this disclosure is not limited to the specific details of the above embodiments. Within the scope of the technical concept of this disclosure, various simple modifications can be made to the technical solutions of this disclosure, and these simple modifications all fall within the protection scope of this disclosure.

[0139] It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. In order to avoid unnecessary repetition, this disclosure will not describe the various possible combinations separately.

[0140] Furthermore, various different embodiments of this disclosure can be combined in any way, as long as they do not violate the spirit of this disclosure, they should also be regarded as the content disclosed in this disclosure.

Claims

1. A material theft prevention method based on a material type fingerprint texture pattern, characterized by, The method comprises the following steps: scanning the materials carried by the vehicle at the starting position and the arrival position of the material transportation by sensors respectively to obtain the three-dimensional shape information of the materials at the corresponding positions; preprocessing the three-dimensional shape information of the materials to obtain the contour line texture map; analyzing and comparing the contour line texture maps at the starting position and the arrival position, and performing similarity calculation; executing the preset similarity logic on the similarity calculation result, or analyzing and comparing the similarity calculation result with the similarity threshold to determine whether the material shape has changed abnormally or the material has been stolen during the transportation; before scanning the materials carried by the vehicle at the starting position and the arrival position of the material transportation by sensors respectively to obtain the three-dimensional shape information of the materials at the corresponding positions, the method further comprises the following steps: scanning the vehicle carrying the materials at the starting position and the arrival position by a camera to obtain the first vehicle feature information of the vehicle; after scanning the materials carried by the vehicle at the starting position and the arrival position of the material transportation by sensors respectively to obtain the three-dimensional shape information of the materials at the corresponding positions, the method further comprises the following steps: grouping the three-dimensional shape information of the materials according to the first vehicle feature information; obtaining at least one group of the three-dimensional shape information of the materials corresponding to the first vehicle feature information.

2. The material theft prevention method based on a material fingerprint texture pattern according to claim 1, wherein, The analyzing and comparing the contour line texture maps at the starting position and the arrival position, and performing similarity calculation comprises the following steps: analyzing the contour line texture map at the starting position to obtain the center of gravity or center position information of at least one group of concentric closed curves in the contour line texture map; analyzing the contour line texture map at the arrival position to obtain the center of gravity or center position information of at least one group of concentric closed curves in the contour line texture map; comparing the center of gravity or center position information change vectors of at least one group of concentric closed curves at the starting position and the arrival position; performing similarity calculation according to the center of gravity or center position information change vectors of at least one group of concentric closed curves.

3. The material theft prevention method based on a material fingerprint texture pattern according to claim 1, wherein, The analyzing and comparing the contour line texture maps at the starting position and the arrival position, and performing similarity calculation comprises the following steps: analyzing the contour line texture map at the starting position to obtain the center of gravity or center position information of each concentric closed curve in each group of concentric closed curves in the contour line texture map; analyzing the contour line texture map at the arrival position to obtain the center of gravity or center position information of each concentric closed curve in each group of concentric closed curves in the contour line texture map; comparing the center of gravity or center position change vectors of each concentric closed curve in each group of concentric closed curves at the starting position and the arrival position; performing similarity calculation according to the center of gravity or center position change vectors of each concentric closed curve in each group of concentric closed curves.

4. The material theft prevention method based on a material fingerprint texture pattern according to claim 1, wherein, The analyzing and comparing the contour line texture maps at the starting position and the arrival position, and performing similarity calculation comprises the following steps: analyzing the contour line texture map at the starting position to obtain the center of gravity or center position information of each group of concentric closed curves in the multiple groups of concentric closed curves in the contour line texture map; analyzing the contour texture at the arrival position to obtain the center position information of each of the plurality of sets of concentric closed curves in the contour texture; comparing the change vector of the center position of each of the plurality of sets of concentric closed curves at the starting position and the arrival position; performing similarity calculation according to the change vector of the center position of each of the plurality of sets of concentric closed curves.

5. The material theft prevention method based on a material fingerprint texture pattern according to claim 1, wherein, The analysis and comparison of the contour texture at the starting position and the arrival position and the similarity calculation include the following steps: obtaining the center position information of the plurality of sets of concentric closed curves in the contour texture at the starting position and the arrival position; calculating the relative position change vector of the center of each two sets of concentric closed curves from the starting position to the arrival position; performing similarity calculation according to the relative position change vector of the center of each two sets of concentric closed curves.

6. The material theft prevention method based on the material fingerprint texture pattern according to any one of claims 2-5, characterized in that, The analysis and comparison of the contour texture at the starting position and the arrival position and the similarity calculation further include the following steps: obtaining the plurality of sets of concentric closed curves in the contour texture at the starting position and the arrival position; analyzing and comparing the distance change vector between the adjacent two concentric closed curves in each set of concentric closed curves, and performing similarity calculation according to the distance change vector.

7. The material theft prevention method based on a material fingerprint texture pattern according to claim 1, wherein, The method further includes the following steps: obtaining the material quality information and / or the material volume information in the material three-dimensional morphological information; analyzing and comparing the material quality information and / or the material volume information at the starting position and the arrival position to determine whether the material is stolen during transportation.

8. The material theft prevention method based on a material fingerprint texture pattern according to claim 1, wherein, The method further includes the following steps: if it is determined that the material has changed during transportation or the degree of change of the material during transportation exceeds a set threshold, a material theft warning is prompted.

9. The stock theft prevention method based on a material type fingerprint texture pattern according to claim 1, wherein, The method further includes the following steps: scanning the vehicle transporting the material at the starting position and the arrival position by the sensor to obtain second vehicle feature information of the vehicle; performing preprocessing on the material three-dimensional morphological information and the second vehicle feature information to obtain a contour texture of the vehicle and the material as a whole.

10. The material theft prevention method based on a material fingerprint texture pattern according to claim 9, wherein, The preprocessing of the material three-dimensional morphological information and the second vehicle feature information to obtain a contour texture of the vehicle and the material as a whole includes the following steps: using the vehicle frame boundary information in the second vehicle feature information as the material region range; establishing a reference coordinate system according to the vehicle frame boundary information, and converting the point cloud data information in the material three-dimensional morphological information corresponding to the starting position and the arrival position of the material transportation to the reference coordinate system; performing preprocessing on the converted point cloud data information in the material three-dimensional morphological information in combination with the material region range to obtain a contour texture of the vehicle and the material as a whole.

Images