Cargo modeling method and device

Abstract

The invention discloses a cargo modeling method and device, and the method comprises the steps: carrying out the marking of a cargo, and obtaining the label information of the cargo; the label information comprises type information and weight information; detecting the goods to obtain a goods point cloud picture; preprocessing the cargo point cloud atlas to obtain a preprocessed point cloud atlas; performing key point extraction on the preprocessed point cloud picture to obtain a key point sequence; and integrating the cargo point cloud picture, the cargo label information and the key point sequence to obtain a cargo 3D model. According to the method, the position of the goods in the packaging container can be visually seen through the 3D perspective view, the volume utilization rate of the packaging container can be improved through the goods 3D model, the gravity center of the goods after boxing can be lowered, and the safety of goods stacking storage and transportation is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of logistics and distribution, and in particular to a cargo modeling method and device. Background Art

[0002] In the logistics field, determining cargo status is crucial during the packing of cargo into box containers and loading of cargo onto transport vehicles. Accurate and timely cargo status identification can help managers improve vehicle management efficiency and reduce transportation costs. Currently, cargo status identification is primarily performed manually, with drivers or staff manually counting and reporting the actual cargo status. However, this method suffers from information lags and significantly increases labor costs.

[0003] Cargo modeling, a crucial component of logistics system modeling and digital warehouse management, relies heavily on the development of 3D reconstruction and digital modeling technologies. These technologies enable the accurate and efficient creation of 3D cargo models, providing powerful support for the simulation, optimization, and visualization of logistics systems. Summary of the Invention

[0004] The technical problem to be solved by this invention is to provide a cargo modeling method and device that automatically collects the cargo's dimensions and weight through a 3D digital model acquisition platform to generate a 3D digital model. The 3D perspective view allows for intuitive visualization of the cargo's position within the packaging container, improving the container's volume utilization and lowering the cargo's center of gravity after packaging, thereby increasing safety during stacking, storage, and transportation.

[0005] In order to solve the above technical problems, a first aspect of an embodiment of the present invention discloses a cargo modeling method, the method comprising:

[0006] S1, labeling the goods to obtain goods label information; the label information includes type information and weight information;

[0007] S2, inspecting the goods to obtain a point cloud map of the goods;

[0008] S3: Process the cargo point cloud image to obtain a 3D model of the cargo.

[0009] As an optional implementation manner, in the first aspect of the embodiment of the present invention, processing the cargo point cloud image to obtain a cargo 3D model includes:

[0010] S31, preprocessing the cargo point cloud map to obtain a preprocessed point cloud map;

[0011] S32, extracting key points from the preprocessed point cloud image to obtain a key point sequence;

[0012] S33: Integrate the cargo point cloud image, the cargo label information, and the key point sequence to obtain a cargo 3D model.

[0013] As an optional implementation manner, in the first aspect of the embodiment of the present invention, preprocessing the cargo point cloud map to obtain a preprocessed point cloud map includes:

[0014] S311, performing smoothing filtering on the cargo point cloud image to obtain a first point cloud image;

[0015] S312, removing outliers from the first point cloud image to obtain a second point cloud image;

[0016] S313, downsampling the second point cloud image to obtain a third point cloud image;

[0017] S314: Perform data registration on the third point cloud image to obtain a preprocessed point cloud image.

[0018] As an optional implementation manner, in the first aspect of the embodiment of the present invention, performing smoothing filtering on the cargo point cloud map to obtain the first point cloud map includes:

[0019] Using a smoothing filter model, smoothing the cargo point cloud image is performed to obtain a first point cloud image;

[0020] The smoothing filter model expression is:

[0021]

[0022] Among them, n is the number of point clouds, w(r i ) is the point (x i ,y i ,z i ), F is the first point cloud, p(x,y)=[p1(x,y),p2(x,y),…,p m (x,y)] is the basis function, T represents the transpose, m is the number of basis functions, A(x,y)=[a1(x,y),a2(x,y),…,a m (x,y)] is the filter coefficient.

[0023] As an optional implementation manner, in the first aspect of the embodiment of the present invention, removing outliers from the first point cloud image to obtain a second point cloud image includes:

[0024] S3121, transforming the first point cloud image to obtain a two-dimensional image;

[0025] S3122, performing binarization processing on the two-dimensional image to obtain a binarized image;

[0026] S3123, filtering the binary image to obtain a filtered image;

[0027] S3124, performing edge extraction on the filtered image to obtain an edge image;

[0028] S3125, performing transformation processing on the edge image to obtain a second point cloud map.

[0029] As an optional implementation, in the first aspect of the embodiment of the present application, the data down-sampling of the second point cloud map to obtain a third point cloud map comprises:

[0030] S3131, the second point cloud map is divided to obtain M voxel grids, M being a positive integer;

[0031] S3132, the M voxel grids are respectively subjected to cluster analysis to obtain M voxel grid centers;

[0032] S3133, the M voxel grid centers are fitted to obtain the third point cloud map.

[0033] As an optional implementation, in the first aspect of the embodiment of the present application, the key point extraction of the pre-processed point cloud map to obtain a key point sequence comprises:

[0034] S321, performing feature calculation on the pre-processed point cloud map to obtain a feature value;

[0035] S322, using a pre-set feature threshold to screen the feature value to obtain an optimized feature value;

[0036] S323, clustering the optimized feature value to obtain a first key point sequence;

[0037] S324, correcting the first key point sequence to obtain the key point sequence.

[0038] The second aspect of the embodiment of the present application discloses a cargo modeling device, which comprises:

[0039] A cargo labeling module is configured to label the cargo to obtain cargo label information; the label information comprises category information and weight information;

[0040] A cargo detection module is configured to detect the cargo to obtain a cargo point cloud map;

[0041] A cargo 3D model construction module is configured to process the cargo point cloud map to obtain a cargo 3D model.

[0042] As an optional implementation manner, in the second aspect of the embodiment of the present invention, processing the cargo point cloud image to obtain a cargo 3D model includes:

[0043] S31, preprocessing the cargo point cloud map to obtain a preprocessed point cloud map;

[0044] S32, extracting key points from the preprocessed point cloud image to obtain a key point sequence;

[0045] S33: Integrate the cargo point cloud image, the cargo label information, and the key point sequence to obtain a cargo 3D model.

[0046] As an optional implementation manner, in the second aspect of the embodiment of the present invention, preprocessing the cargo point cloud map to obtain a preprocessed point cloud map includes:

[0047] S311, performing smoothing filtering on the cargo point cloud image to obtain a first point cloud image;

[0048] S312, removing outliers from the first point cloud image to obtain a second point cloud image;

[0049] S313, downsampling the second point cloud image to obtain a third point cloud image;

[0050] S314: Perform data registration on the third point cloud image to obtain a preprocessed point cloud image.

[0051] As an optional implementation manner, in the second aspect of the embodiment of the present invention, performing smoothing filtering on the cargo point cloud map to obtain the first point cloud map includes:

[0052] Using a smoothing filter model, smoothing the cargo point cloud image is performed to obtain a first point cloud image;

[0053] The smoothing filter model expression is:

[0054]

[0055] Among them, n is the number of point clouds, w(r i ) is the point (x i ,y i ,z i ), F is the first point cloud, p(x,y)=[p1(x,y),p2(x,y),…,p m (x,y)] is the basis function, T represents the transpose, m is the number of basis functions, A(x,y)=[a1(x,y),a2(x,y),…,a m (x,y)] is the filter coefficient.

[0056] As an optional implementation manner, in the second aspect of the embodiment of the present invention, removing outliers from the first point cloud image to obtain a second point cloud image includes:

[0057] S3121, transforming the first point cloud image to obtain a two-dimensional image;

[0058] S3122, performing binarization processing on the two-dimensional image to obtain a binarized image;

[0059] S3123, filtering the binarized image to obtain a filtered image;

[0060] S3124, performing edge extraction on the filtered image to obtain an edge image;

[0061] S3125: Perform transformation processing on the edge image to obtain a second point cloud image.

[0062] As an optional implementation manner, in the second aspect of the embodiment of the present invention, downsampling the second point cloud image to obtain a third point cloud image includes:

[0063] S3131, dividing the second point cloud image into M voxel grids, where M is a positive integer;

[0064] S3132, performing cluster analysis on the M voxel grids to obtain the centroids of the M voxel grids;

[0065] S3133 , fitting the centroids of the M voxel grids to obtain a third point cloud image.

[0066] As an optional implementation manner, in the second aspect of the embodiment of the present invention, extracting key points from the preprocessed point cloud image to obtain a key point sequence includes:

[0067] S321, performing feature calculation on the pre-processed point cloud image to obtain feature values;

[0068] S322, screening the characteristic value using a preset characteristic threshold to obtain an optimized characteristic value;

[0069] S323, clustering the optimized eigenvalues ​​to obtain a first key point sequence;

[0070] S324: Correct the first key point sequence to obtain a key point sequence.

[0071] A third aspect of the present invention discloses another cargo modeling device, comprising:

[0072] a memory storing executable program code;

[0073] a processor coupled to the memory;

[0074] The processor calls the executable program code stored in the memory to execute part or all of the steps in the cargo modeling method disclosed in the first aspect of the embodiment of the present invention.

[0075] A fourth aspect of the present invention discloses a computer-storable medium storing computer instructions. When the computer instructions are called, they are used to execute some or all of the steps in the cargo modeling method disclosed in the first aspect of the embodiment of the present invention.

[0076] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0077] The present invention uses a 3D digital model acquisition platform to collect cargo type and weight information; inspects the cargo to generate a cargo point cloud map; and processes the cargo point cloud map to generate a 3D cargo model. The 3D cargo model allows for intuitive visualization of the cargo's position within the packaging container, improving the container's volumetric utilization. It also lowers the center of gravity of the packed cargo, increasing safety during stacking, storage, and transportation. The method can be used in the logistics and distribution sector, both for packing cargo into box-type containers and for loading cargo onto transport vehicles. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0079] Figure 1 This is a flow chart of a cargo modeling method disclosed in an embodiment of the present invention;

[0080] Figure 2 is a schematic diagram of a filter disclosed in an embodiment of the present invention;

[0081] Figure 3 This is a flow chart of a cargo modeling method disclosed in an embodiment of the present invention;

[0082] Figure 4 This is a structural diagram of a cargo modeling device disclosed in an embodiment of the present invention;

[0083] Figure 5 It is a structural schematic diagram of another cargo modeling device disclosed in an embodiment of the present invention. DETAILED DESCRIPTION

[0084] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0085] The terms "first," "second," and so on, in the description and claims of the present invention and the accompanying drawings are used to distinguish between different objects, not to describe a specific order. Furthermore, the terms "including," "having," and any variations thereof, are intended to cover non-exclusive inclusions. For example, a process, method, apparatus, product, or device comprising a series of steps or elements is not limited to the listed steps or elements but may optionally include steps or elements not listed, or may optionally include other steps or elements inherent to the process, method, product, or device.

[0086] References herein to "embodiments" mean that a particular feature, structure, or characteristic described in connection with the embodiments may be included in at least one embodiment of the present invention. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor does it constitute a separate or alternative embodiment that is mutually exclusive of other embodiments. It is understood, both explicitly and implicitly, by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0087] The present invention discloses a cargo modeling method and device, which includes: marking cargo to obtain cargo label information; the label information includes type information and weight information; inspecting the cargo to obtain a cargo point cloud map; preprocessing the cargo point cloud map to obtain a preprocessed point cloud map; extracting key points from the preprocessed point cloud map to obtain a key point sequence; integrating the cargo point cloud map, the cargo label information, and the key point sequence to obtain a cargo 3D model. The present invention utilizes a 3D perspective view to intuitively see the position of the cargo in the packaging container. The use of the cargo 3D model is conducive to improving the volume utilization rate of the packaging container, and can also reduce the center of gravity of the cargo after packaging, thereby increasing the safety of cargo stacking, storage, and transportation. Detailed descriptions are given below.

[0088] Example 1

[0089] See also Figure 1 , Figure 1 This is a flow chart of a cargo modeling method disclosed in an embodiment of the present invention. Figure 1 The cargo modeling method described is applied to the field of logistics and distribution technology, and is not limited in the embodiments of the present invention. Figure 1As shown, the cargo modeling method may include the following operations:

[0090] S1, labeling the goods to obtain goods label information; the label information includes type information and weight information;

[0091] S2, inspecting the goods to obtain a point cloud map of the goods;

[0092] Optionally, the cargo is detected using methods such as visual SLAM to obtain a cargo point cloud map;

[0093] S3: Process the cargo point cloud image to obtain a 3D model of the cargo.

[0094] Optionally, processing the cargo point cloud image to obtain a cargo 3D model includes:

[0095] S31, preprocessing the cargo point cloud map to obtain a preprocessed point cloud map;

[0096] S32, extracting key points from the preprocessed point cloud image to obtain a key point sequence;

[0097] S33: Integrate the cargo point cloud image, the cargo label information, and the key point sequence to obtain a cargo 3D model.

[0098] Optionally, preprocessing the cargo point cloud map to obtain a preprocessed point cloud map includes:

[0099] S311, performing smoothing filtering on the cargo point cloud image to obtain a first point cloud image;

[0100] S312, removing outliers from the first point cloud image to obtain a second point cloud image;

[0101] S313, downsampling the second point cloud image to obtain a third point cloud image;

[0102] S314: Perform data registration on the third point cloud image to obtain a preprocessed point cloud image.

[0103] Optionally, performing smoothing filtering on the cargo point cloud image to obtain a first point cloud image includes:

[0104] Using a smoothing filter model, smoothing the cargo point cloud image is performed to obtain a first point cloud image;

[0105] The smoothing filter model expression is:

[0106]

[0107] Among them, n is the number of point clouds, w(r i) is the point (x i ,y i ,z i ), F is the first point cloud, p(x,y)=[p1(x,y),p2(x,y),…,p m (x,y)] is the basis function, T represents the transpose, m is the number of basis functions, A(x,y)=[a1(x,y),a2(x,y),…,a m (x,y)] is the filter coefficient.

[0108] Optionally, removing outliers from the first point cloud image to obtain a second point cloud image includes:

[0109] S3121, transforming the first point cloud image to obtain a two-dimensional image;

[0110] The specific calculation formula is:

[0111] u=[M1M2] T P

[0112] Where u is a two-dimensional image, P is the first point cloud, and u = (x, y) T , P=(X,Y,Z) T , T represents transpose, M1 and M2 are transformation matrices, which are set by specific experiments. In this embodiment,

[0113] S3122, performing binarization processing on the two-dimensional image to obtain a binarized image;

[0114] S3123, filtering the binarized image to obtain a filtered image, including:

[0115] Perform time-frequency analysis on the binary image to obtain time-frequency data information X s , For X s The kth data in , k=1,2,…,K, K is the data length;

[0116] Preset reference time-frequency data information X c , For X c The kth data in ;

[0117] X s and X c Composition data pair

[0118] Sample a data from P:

[0119] Set the time step: t~Uniform({1,...,T}), the uniform() function is used to generate random numbers, and T is the upper limit of the step size;

[0120] Get the denoised data information of step t for:

[0121]

[0122] in, α t =1-β t , β t ∈(0,1), represents a normal distribution between 0 and 1. Use the trained filter to t , The result obtained by processing;

[0123] Filters such as Figure 2 shown.

[0124] S3124, performing edge extraction on the filtered image to obtain an edge image;

[0125] Edge extraction is achieved using the Laplacian operator;

[0126] S3125, transforming the edge image to obtain a second point cloud image, including:

[0127] Assume that the edge image is X input , using PointNet++ network to input Processing to obtain the first feature information Where N0 is the number of points in the point cloud, C1 is the feature dimension of the first feature information, and the global feature is obtained by extracting the last layer of features.

[0128] Process the first feature information to obtain the second feature information

[0129]

[0130] Among them, C2 is the feature dimension of the second feature information, ρ(·) is the copy operation, and cat(·) is the splicing operation; the global feature and the first feature information are processed to obtain the third feature

[0131]

[0132] in, For the multi-layer perceptron; for the third feature, global feature and X input Processing is performed to obtain the fourth feature F missing :

[0133]

[0134] F know and F missing Copy A1 and A2 times respectively to get F c , A1 and A2 are preset integers, and F is transformed into c Projection is extended to high-dimensional space to obtain a point set with N1 points N1=(A1+A2)×N0, with X coarse With each point in the image as the center, A3 points are generated in the local neighborhood of each center point through the folding operation, and a second point cloud containing N2 dense complete shape points is obtained, where N2 = N1 × (A3 + 1), and A3 is a preset integer.

[0135] Optionally, downsampling the second point cloud image to obtain a third point cloud image includes:

[0136] S3131, dividing the second point cloud image into M voxel grids, where M is a positive integer;

[0137] S3132, performing cluster analysis on the M voxel grids to obtain the centroids of the M voxel grids;

[0138] S3133 , fitting the centroids of the M voxel grids to obtain a third point cloud image.

[0139] Optionally, extracting key points from the preprocessed point cloud image to obtain a key point sequence includes:

[0140] S321, performing feature calculation on the pre-processed point cloud image to obtain feature values;

[0141] The calculation formula of the eigenvalue is:

[0142]

[0143] Where K is the characteristic of point P, the function expression of point P is y=f(x), the slope of the tangent line at point P is y′, y" is the second-order derivative of y, and y′ is the first-order derivative of y. is a parametric equation expression, the superscript ' represents the first-order derivative, the superscript " represents the second-order derivative, and t is the time variable;

[0144] S322, screening the characteristic value using a preset characteristic threshold to obtain an optimized characteristic value;

[0145] The feature threshold is set by experiments, and the point cloud is segmented based on this threshold to obtain the optimized feature value (point group with features).

[0146] S323, clustering the optimized eigenvalues ​​to obtain a first key point sequence;

[0147] That is to find the density center of the corresponding point cloud, so it is done through clustering method;

[0148] 1) First, k initial cluster centers need to be selected;

[0149] 2) Then calculate the distance from each initialized cluster center to each data point, and assign the data point to the cluster center with the closest distance until all points are divided into k data sets;

[0150] 3) Then recalculate the mean of the data points in each data set and update the mean as the new cluster center;

[0151] 4) Finally, calculate the distance between each data point and the new k initialized cluster centers and re-divide;

[0152] 5) Continue to traverse new cluster centers and iterate until the data is stable and the data sets do not change. The cluster centers of the processed data sets are the points with high point cloud density, and k cluster centers are obtained;

[0153] S324: Correct the first key point sequence to obtain a key point sequence.

[0154] By evaluating the Euclidean distance between the cluster center (the first key point sequence) and the original point cloud, when a point has the smallest distance from the cluster center, it is the corresponding feature point (that is, the expression of the key point in the original point cloud).

[0155] As can be seen, the present invention uses a 3D digital model acquisition platform to collect cargo type and weight information; inspects the cargo to obtain a cargo point cloud map; and processes the cargo point cloud map to obtain a 3D cargo model. The 3D cargo model allows for intuitive visualization of the cargo's position within the packaging container, which helps improve the container's volumetric utilization. It also lowers the center of gravity of the packed cargo, increasing safety during stacking, storage, and transportation. The present method can be used in the logistics and distribution field, both for packing cargo into box-type containers and for loading cargo onto transport vehicles.

[0156] Example 2

[0157] See also Figure 3 , Figure 3 This is a flow chart of another cargo modeling method disclosed in an embodiment of the present invention. Figure 3The cargo modeling method described is applied to the field of logistics and distribution technology, and is not limited in the embodiments of the present invention. Figure 3 As shown, the cargo modeling method may include the following operations:

[0158] The 3D digital model acquisition platform automatically captures the cargo's dimensions and weight, generating a 3D digital model. The 3D perspective view allows for a visual representation of the cargo's position within the packaging container, improving container volume utilization and lowering the cargo's center of gravity after packing, increasing safety during stacking, storage, and transportation.

[0159] Step 1: Place the goods on the 3D digital model acquisition platform. The goods should be placed in the center of the platform as much as possible. If the size of the goods exceeds the platform, replace the scanning platform with a larger size.

[0160] Step 2: Scan the cargo shape, generate a 3D point cloud map of the cargo, and automatically collect the cargo's dimensions and weight;

[0161] Step 3: Turn the goods over, scan the shape of the goods again and generate a 3D point cloud map of the goods, while automatically collecting the dimensions of the goods;

[0162] Step 4: Compare the two scan results and make corrections to generate a 3D digital model of the cargo with dimensions and weight parameters.

[0163] Example 3

[0164] See also Figure 4 , Figure 4 This is a structural diagram of a cargo modeling device disclosed in an embodiment of the present invention. Figure 4 The cargo modeling device described is applied to the field of logistics and distribution technology, and the embodiments of the present invention are not limited thereto. Figure 4 As shown, the cargo modeling device may include the following operations:

[0165] S301, a cargo labeling module is used to label cargo and obtain cargo label information; the label information includes type information and weight information;

[0166] S302, a cargo detection module, configured to detect the cargo and obtain a cargo point cloud map;

[0167] S303, a cargo 3D model building module is used to process the cargo point cloud image to obtain a cargo 3D model.

[0168] Example 4

[0169] See also Figure 5 , Figure 5 This is a schematic diagram of the structure of another cargo modeling device disclosed in an embodiment of the present invention. Figure 5 The cargo modeling device described is applied to the field of logistics and distribution technology, and the embodiments of the present invention are not limited thereto. Figure 5 As shown, the cargo modeling device may include the following operations:

[0170] A memory 401 storing executable program code;

[0171] a processor 402 coupled to the memory 401;

[0172] The processor 402 calls the executable program code stored in the memory 401 to execute the steps in the cargo modeling method described in the first and second embodiments.

[0173] Example 5

[0174] An embodiment of the present invention discloses a computer-readable storage medium storing a computer program for electronic data exchange, wherein the computer program enables a computer to execute the steps of the cargo modeling method described in the first and second embodiments.

[0175] The device embodiments described above are merely illustrative. Modules described as separate components may or may not be physically separate, and components shown as modules may or may not be physical modules, i.e., they may be located in one place or distributed across multiple network modules. Some or all of these modules may be selected based on actual needs to achieve the objectives of the present embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0176] Through the detailed description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus the necessary general hardware platform, or of course, by means of hardware. Based on this understanding, the above technical solution, in essence, or the portion that contributes to the prior art, can be embodied in the form of a software product, which can be stored in a computer-readable storage medium, including a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium capable of carrying or storing data.

[0177] Finally, it should be noted that the cargo modeling method and device disclosed in the embodiments of the present invention are only preferred embodiments of the present invention and are only used to illustrate the technical solutions of the present invention, rather than to limit them. Although the present invention has been described in detail with reference to the aforementioned embodiments, it should be understood by those skilled in the art that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A cargo modeling method, characterized in that: The method comprises: S1, labeling the goods to obtain goods label information; the label information includes type information and weight information; S2, inspecting the goods to obtain a point cloud map of the goods; S3: Process the cargo point cloud image to obtain a 3D model of the cargo.

2. The cargo modeling method according to claim 1, characterized in that: The processing of the cargo point cloud image to obtain a cargo 3D model includes: S31, preprocessing the cargo point cloud map to obtain a preprocessed point cloud map; S32, extracting key points from the preprocessed point cloud image to obtain a key point sequence; S33: Integrate the cargo point cloud image, the cargo label information, and the key point sequence to obtain a cargo 3D model.

3. The cargo modeling method according to claim 2, characterized in that: The preprocessing of the cargo point cloud map to obtain a preprocessed point cloud map includes: S311, performing smoothing filtering on the cargo point cloud image to obtain a first point cloud image; S312, removing outliers from the first point cloud image to obtain a second point cloud image; S313, downsampling the second point cloud image to obtain a third point cloud image; S314: Perform data registration on the third point cloud image to obtain a preprocessed point cloud image.

4. The cargo modeling method according to claim 3, characterized in that: The step of performing smoothing filtering on the cargo point cloud image to obtain a first point cloud image includes: Using a smoothing filter model, smoothing the cargo point cloud image is performed to obtain a first point cloud image; The smoothing filter model expression is: Among them, n is the number of point clouds, w(r i ) is the point (x i ,y i ,z i ), F is the first point cloud, p(x,y)=[p1(x,y),p2(x,y),…,p m (x,y)] is the basis function, T represents the transpose, m is the number of basis functions, A(x,y)=[a1(x,y),a2(x,y),…,a m (x,y)] is the filter coefficient.

5. The cargo modeling method according to claim 3, characterized in that: The step of removing outliers from the first point cloud image to obtain a second point cloud image includes: S3121, transforming the first point cloud image to obtain a two-dimensional image; S3122, performing binarization processing on the two-dimensional image to obtain a binarized image; S3123, filtering the binarized image to obtain a filtered image; S3124, performing edge extraction on the filtered image to obtain an edge image; S3125: Perform transformation processing on the edge image to obtain a second point cloud image.

6. The cargo modeling method according to claim 3, characterized in that: The downsampling of the second point cloud image to obtain a third point cloud image includes: S3131, dividing the second point cloud image into M voxel grids, where M is a positive integer; S3132, performing cluster analysis on the M voxel grids to obtain the centroids of the M voxel grids; S3133 , fitting the centroids of the M voxel grids to obtain a third point cloud image.

7. The cargo modeling method according to claim 2, characterized in that: The step of extracting key points from the pre-processed point cloud image to obtain a key point sequence includes: S321, performing feature calculation on the pre-processed point cloud image to obtain feature values; S322, screening the characteristic values ​​using a preset characteristic threshold to obtain an optimized characteristic value; S323, clustering the optimized eigenvalues ​​to obtain a first key point sequence; S324: Correct the first key point sequence to obtain a key point sequence.

8. A cargo modeling device, characterized in that: The device comprises: The cargo labeling module is used to label the cargo and obtain cargo label information; the label information includes type information and weight information; A cargo detection module, configured to detect the cargo and obtain a cargo point cloud map; The cargo 3D model building module is used to process the cargo point cloud image to obtain a cargo 3D model.

9. A cargo modeling device, characterized in that: The device comprises: a memory storing executable program code; a processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the cargo modeling method according to any one of claims 1 to 7.

10. A computer storable medium, characterized in that The computer storable medium stores computer instructions, and when the computer instructions are called, they are used to execute the cargo modeling method according to any one of claims 1 to 7.

Images