Container cargo loading contour detection method, device, equipment and medium

Abstract

The present application relates to the technical field of computer vision, and provides a kind of container plate cargo loading profile detection method, device, equipment and medium, method includes: obtaining the spatial coordinate information of cargo packaging surface, obtain a plurality of test points;According to each test point, in combination with the reference plate type model of container plate obtained in advance, determine the unit topological relation of each test point and reference plate type model;Wherein, reference plate type model is obtained by equidistance inflation based on the three-dimensional reference model corresponding to container plate type in advance, and unit topological relation is used to represent the spatial position relation between corresponding test point and basic geometric unit in reference plate type model;According to unit topological relation, when test point is located outside the space envelope, obtain the out-of-limit point;According to each out-of-limit point, determine the maximum spatial distance of out-of-limit point and each surface of reference plate type model, obtain the profile detection result.The present application can accurately deal with various complex cargo forms, realize the digital tracing of detection whole process, and improve the detection efficiency.

Description

Technical Field

[0001] This invention relates to the field of computer vision technology, and in particular to a method, apparatus, equipment and medium for detecting the outline of cargo loading on pallets. Background Technology

[0002] As a crucial pillar of the modern logistics system, the safety and efficiency of cargo transportation have always been the primary goals of the air transport industry. As core equipment in air freight, the compliance of pallet loading profiles not only affects the safety of individual shipments but is also a key factor influencing the operational efficiency of the entire air logistics system. According to the airworthiness standards of the International Air Transport Association (IATA), pallets must strictly comply with the outline restrictions of aircraft cargo holds; any dimensional deviation can lead to significant safety hazards such as loading failures and damage to the cargo hold structure. Therefore, pallet cargo loading profile inspection, as an indispensable quality control link in the air freight process, through systematic dimensional inspection and airworthiness assessment, not only ensures transportation safety but also significantly improves cargo hold space utilization, which is of strategic significance for optimizing the overall operational efficiency of air logistics.

[0003] Currently, the inspection of palletized cargo loading contours in the air cargo sector is still dominated by traditional manual inspection methods, mainly relying on visual inspection by operators and measurement with simple measuring tools (such as calipers).

[0004] However, this manual inspection method suffers from limitations in comprehensive coverage, low efficiency, inefficient data management, poor standard adaptability, and increasingly prominent unavoidable human error. Furthermore, it struggles to address the varying standardized unit load device (ULD) profiles across different airlines, and lacks effective countermeasures for the unique "compression rebound" phenomenon of e-commerce goods. More seriously, the frequent work at heights during inspection poses significant safety hazards, threatening the personal safety of personnel. Summary of the Invention

[0005] This invention provides a method, apparatus, equipment, and medium for detecting the loading contour of containerized cargo, which solves the problems of low efficiency and large errors in manual inspection in the prior art, realizes digital traceability of the entire inspection process, and improves inspection efficiency.

[0006] The application provides a container plate cargo loading contour detection method, comprising: obtaining spatial coordinate information of a cargo packaging surface to obtain a plurality of to-be-detected points; determining a unit topological relationship between each to-be-detected point and a reference plate model based on the to-be-detected points and the reference plate model obtained in advance; wherein the reference plate model is obtained by equidistantly expanding a three-dimensional reference model corresponding to the container plate type in advance, and the unit topological relationship is used to represent the spatial position relationship between the to-be-detected point and a basic geometric unit in the reference plate model; when the to-be-detected point is located outside the spatial envelope of the reference plate model, obtaining an out-of-limit point based on the unit topological relationship; and obtaining a contour detection result based on the spatial maximum distance between each out-of-limit point and each surface of the reference plate model.

[0007] According to the container plate cargo loading contour detection method provided by the application, the spatial maximum distance between each out-of-limit point and each surface of the reference plate model is determined to obtain a contour detection result, comprising: performing density-based clustering based on all out-of-limit points to obtain out-of-limit region clusters; extracting and fitting a main plane of each out-of-limit region cluster to obtain a plurality of plane clusters with the same distance representation; for each plane cluster, traversing the points in the plane cluster to determine the spatial maximum distance and the normal angle between each point in the plane cluster and each surface of the reference plate model; and obtaining the contour detection result based on the spatial maximum distance and the normal angle between each point in the plane cluster and each surface of the reference plate model.

[0008] According to the container plate cargo loading contour detection method provided by the application, the spatial maximum distance and the normal angle between each point in the plane cluster and each surface of the reference plate model are determined to obtain a contour detection result, comprising: selecting the smallest spatial maximum distance as the characteristic distance of the plane cluster based on the spatial maximum distance between each point in the plane cluster and each surface of the reference plate model; determining a to-be-mapped orientation based on the normal angle between the point corresponding to the smallest spatial maximum distance and each surface of the reference plate model, combining a preset mapping orientation rule, and mapping the corresponding point to the surface of the reference plate model according to the to-be-mapped orientation to obtain the position orientation positioning information of the corresponding plane cluster; wherein the preset mapping orientation rule is configured in advance according to the value range of different normal angles and the corresponding mapping orientation; and obtaining the contour detection result based on the characteristic distance of the plane cluster and the position orientation positioning information of the plane cluster.

[0009] The application provides a container plate cargo loading contour detection method, before determining the unit topological relation between each to-be-detected point and the reference plate type model according to each to-be-detected point and the reference plate type model acquired in advance, comprising the following steps of: calling a three-dimensional reference model of the corresponding container plate in a preset parameter library according to the container plate type determined in advance; wherein the preset parameter library comprises three-dimensional reference models of container plates of various plate types, and the three-dimensional reference model is constructed in advance based on the corresponding plate type of the container plate; for each topological unit on the surface of the three-dimensional reference model, a space offset vector is generated along the normal direction, and the space offset vector is smoothed and corrected; and the reference plate type model is obtained by using equidistance inflation according to the corrected space offset vector and the three-dimensional reference model.

[0010] The application provides a container plate cargo loading contour detection method, and the reference plate type model is obtained by using equidistance inflation according to the corrected space offset vector and the three-dimensional reference model, comprising the following steps of: generating a space envelope by performing equidistance inflation on the three-dimensional reference model according to the corrected space offset vector; performing tetrahedral element division on the internal space of the space envelope, and establishing a space position query index to obtain the reference plate type model.

[0011] The application provides a container plate cargo loading contour detection method, and the space coordinate information of the cargo packaging surface is obtained, comprising the following steps of: acquiring point cloud, a depth map and original space coordinates of container plate cargo packaging; performing interpolation completion on the depth map according to the original space coordinates and the point cloud, and converting the interpolated and completed depth map into three-dimensional point cloud data to obtain fusion information; and obtaining the space coordinate information of the cargo packaging surface according to the fusion information.

[0012] The application provides a container plate cargo loading contour detection method, and the space coordinate information of the cargo packaging surface is obtained according to the fusion information, comprising the following steps of: establishing a point cloud topological network according to the fusion information, and removing topologically abnormal discrete points by using curvature analysis to obtain fusion information after removing the discrete points; and / or extracting an effective data region by using a space region extraction algorithm according to the fusion information or the fusion information after removing the discrete points.

[0013] The application further provides a pallet cargo loading contour detection device, comprising: a data acquisition module, which acquires spatial coordinate information of a cargo packaging surface to obtain a plurality of to-be-detected points; a relationship determination module, which determines a unit topological relationship between each to-be-detected point and a reference pallet model based on the to-be-detected points and a previously acquired reference pallet model; wherein the reference pallet model is obtained by isometric inflation of a three-dimensional reference model corresponding to a pallet type, and the unit topological relationship is used to represent a spatial position relationship between a corresponding to-be-detected point and a basic geometric unit in the reference pallet model; a point position judgment module, which determines an out-of-limit point when a to-be-detected point is located outside a spatial envelope of the reference pallet model based on the unit topological relationship; and a contour detection module, which determines a maximum spatial distance between each out-of-limit point and each surface of the reference pallet model to obtain a contour detection result.

[0014] The application further provides an electronic device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor implements the pallet cargo loading contour detection method according to any one of the above when executing the computer program.

[0015] The application further provides a non-transitory computer-readable storage medium, which stores a computer program, wherein the computer program is executed by a processor to implement the pallet cargo loading contour detection method according to any one of the above.

[0016] The application further provides a computer program product, comprising a computer program, wherein the computer program is executed by a processor to implement the pallet cargo loading contour detection method according to any one of the above.

[0017] The pallet cargo loading contour detection method, device, equipment and medium provided by the application can comprehensively cover the cargo packaging surface by acquiring the spatial coordinate information of the cargo packaging surface, avoid analysis deviation caused by data loss, improve the accuracy of subsequent detection, quickly establish the topological relationship with the to-be-detected points by combining the reference pallet model obtained by isometric inflation, avoid complex geometric calculation, improve the processing efficiency, and then quickly identify the out-of-limit points, avoid the subjectivity and errors of manual detection, do not need manual intervention, can efficiently complete the abnormal detection of large-scale point cloud data, and further calculate the maximum spatial distance between the out-of-limit points and each surface of the reference pallet model to further analyze the distance relationship between the out-of-limit points and the surfaces of the model, ensure the comprehensiveness and accuracy of the contour detection, ensure that various complex cargo forms can be accurately dealt with, flexibly adapt to diversified inspection standard requirements, provide a revolutionary intelligent solution for the aviation cargo industry, realize the digital tracing of the detection whole process, and improve the detection efficiency. BRIEF DESCRIPTION OF DRAWINGS

[0018] In order to more clearly illustrate the technical solutions of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings described below are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative effort based on these drawings.

[0019] Figure 1 is a flowchart of the pallet cargo loading profile detection method provided by the present application.

[0020] Figure 2 is a structural schematic diagram of the pallet cargo loading profile detection device provided by the present application.

[0021] Figure 3 is a structural schematic diagram of the electronic device provided by the present application. DETAILED DESCRIPTION

[0022] In order to make the objects, technical solutions and advantages of the present application clearer, the technical solutions of the present application will be described clearly and completely below with reference to the drawings in the present application. Obviously, the described embodiments are only some embodiments of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present application.

[0023] Figure 1 is a flowchart of the pallet cargo loading profile detection method provided by the present application, as shown in Figure 1 , the method comprises:

[0024] S11, obtaining the spatial coordinate information of the surface of the cargo package, to obtain a plurality of to-be-measured points;

[0025] S12, determining the unit topological relationship between each to-be-measured point and the reference plate type model according to each to-be-measured point and the reference plate type model obtained in advance; wherein the reference plate type model is obtained by equidistant inflation based on the three-dimensional reference model corresponding to the plate type of the pallet in advance, and the unit topological relationship is used to represent the spatial position relationship between the corresponding to-be-measured point and the basic geometric unit in the reference plate type model;

[0026] S13, determining the out-of-limit point when the to-be-measured point is located outside the space envelope of the reference plate type model according to the unit topological relationship;

[0027] S14, determining the maximum spatial distance between each out-of-limit point and each surface of the reference plate type model according to each out-of-limit point, to obtain the profile detection result.

[0028] It should be noted that the step number "S1N" in this specification does not represent the order of the container cargo loading contour detection method. The container cargo loading contour detection method of the present invention is described in detail below.

[0029] Step S11: Obtain the spatial coordinate information of the surface of the goods packaging to obtain multiple test points.

[0030] In this embodiment, obtaining the spatial coordinate information of the cargo packaging surface includes: obtaining the point cloud, depth map, and original spatial coordinates of the pallet cargo packaging; interpolating and completing the depth map based on the original spatial coordinates and point cloud, and converting the interpolated and completed depth map into three-dimensional point cloud data to obtain fused information; and obtaining the spatial coordinate information of the cargo packaging surface based on the fused information.

[0031] It should be added that point clouds and depth maps can be acquired based on single sensors or multi-sensor arrays. When using a multi-sensor array, multiple depth maps will be obtained, independently capturing different perspectives of the scene. This allows for the acquisition of depth information through multi-view stereo vision technology. Therefore, after obtaining the fusion information corresponding to each depth map, the fusion information needs to be stitched together to obtain complete point cloud data, thereby improving data coverage integrity. Additionally, the original spatial coordinates can be acquired simultaneously using infrared coded 3D reconstruction technology. The specific choice can be made according to actual design requirements, and no further limitations are made here.

[0032] It should be noted that before acquiring the corresponding depth map, it is also necessary to perform uniform spatial calibration on the sensors in the sensor array to eliminate imaging deviations caused by camera manufacturing errors, installation deviations, etc., so as to ensure that the acquired image data has high geometric accuracy.

[0033] Furthermore, depth maps can be acquired using sensors such as cameras. If acquisition fails, an empty value will be returned. Based on this, erroneous or failed acquisition data can be flagged and discarded to skip subsequent processing. Similarly, point clouds can be acquired using sensors such as radar. After acquiring the corresponding point cloud, the number of points in the point cloud is obtained through point cloud reconstruction technology. If the number of points is lower than a preset threshold, subsequent processing can be skipped based on a prompt.

[0034] Furthermore, before obtaining the spatial coordinate information of the cargo packaging surface based on the fusion information, the process includes: establishing a point cloud topology network based on the fusion information, and using curvature analysis to identify and remove discrete points with topological anomalies to obtain fusion information after removing discrete points, thereby eliminating discrete interference points; and / or, extracting effective data regions based on the fusion information or the fusion information after removing discrete points, combined with a spatial region extraction algorithm.

[0035] In an optional embodiment, before determining the element topology relationship between each test point and the reference plate model based on each test point and the previously acquired reference plate model of the container plate, the method includes: calling the corresponding three-dimensional reference model of the container plate from a preset parameter library according to the previously determined container plate type; wherein, the preset parameter library includes three-dimensional reference models of container plates of various plate types, and the three-dimensional reference model is constructed based on the corresponding plate type of the container plate; for each topological element on the surface of the three-dimensional reference model, a spatial offset vector is generated along the normal direction, and the spatial offset vector is smoothed to suppress the distortion caused by the surface displacement of the three-dimensional reference model; based on the corrected spatial offset vector and the three-dimensional reference model, the reference plate model is obtained by using equidistant dilation.

[0036] Furthermore, based on the corrected spatial offset vector and the three-dimensional reference model, an equidistant expansion is used to obtain a reference plate model, including: performing equidistant expansion on the three-dimensional reference model based on the corrected spatial offset vector to generate a spatial envelope to adapt to goods packaging of different sizes; dividing the internal space of the spatial envelope into tetrahedral elements and establishing a spatial location query index to obtain the reference plate model.

[0037] Step S12: Based on each test point and the previously acquired reference plate model of the container plate, determine the element topology relationship between each test point and the reference plate model; wherein, the reference plate model is obtained by isometric expansion based on the three-dimensional reference model corresponding to the container plate shape, and the element topology relationship is used to characterize the spatial positional relationship between the corresponding test point and the basic geometric elements in the reference plate model.

[0038] In this embodiment, based on each test point and combined with the previously acquired reference plate model of the container plate, the unit topology relationship between each test point and the reference plate model is determined, including: using a spatial location query index to quickly locate the spatial envelope or adjacent spatial envelope where the test point may be located, and using a preset geometric algorithm to determine whether the test point is located in the spatial envelope or adjacent spatial envelope where it may be located, thereby obtaining the unit topology relationship between the corresponding test point and the reference plate model.

[0039] Step S13: Based on the element topology, when the point to be measured is located outside the spatial envelope of the reference plate model, the out-of-limit point is obtained. It should be noted that the out-of-limit point includes the point's location coordinates and the minimum penetration distance.

[0040] Step S14: Based on each out-of-limit point, determine the maximum spatial distance between the out-of-limit point and each surface of the reference plate model, and obtain the contour detection result.

[0041] In this embodiment, based on each out-of-limit point, the maximum spatial distance between the out-of-limit point and each surface of the reference plate model is determined to obtain the contour detection result. This includes: performing density-based clustering based on all out-of-limit points to separate multiple spatially independent out-of-limit region clusters; extracting and fitting the principal plane of each out-of-limit region cluster to obtain multiple plane clusters with the same distance representation; for each plane cluster, traversing the points within the plane cluster to determine the maximum spatial distance and normal angle between each point within the plane cluster and each surface of the reference plate model; and obtaining the contour detection result based on the maximum spatial distance and normal angle between each point within the plane cluster and each surface of the reference plate model.

[0042] It should be added that, after determining the maximum spatial distance and normal angle between each point within the planar cluster and each surface of the reference plate model, the data for the corresponding point can be represented as {(x, y, z), d}. i , θ i In set form, (x, y, z) represents the point cloud coordinates, and d i θ represents the maximum spatial distance. i Indicates the included angle of the normal.

[0043] Specifically, the contour detection result is obtained based on the maximum spatial distance and normal angle between each point in the planar cluster and each surface of the reference plate model. This includes: selecting the minimum maximum spatial distance as the feature distance of the planar cluster based on the maximum spatial distance between each point in the planar cluster and each surface of the reference plate model; determining the orientation to be mapped based on the normal angle between the point corresponding to the minimum maximum spatial distance and each surface of the reference plate model, combined with a preset mapping orientation rule; and mapping the corresponding point to the surface of the reference plate model according to the orientation to be mapped, thereby obtaining the position orientation positioning information of the corresponding planar cluster. The preset mapping orientation rule is configured in advance based on the different ranges of normal angle values ​​and their corresponding mapping orientations. The contour detection result is obtained based on the feature distance of the planar cluster and the position orientation positioning information of the planar cluster.

[0044] In summary, this invention collects spatial coordinate information of the cargo packaging surface to comprehensively cover the surface, avoiding analytical biases caused by missing data and improving the accuracy of subsequent inspections. Combined with a baseline plate model obtained through equidistant expansion, it quickly establishes a topological relationship with the points to be tested, avoiding complex geometric calculations and improving processing efficiency. This allows for rapid identification of out-of-limit points, avoiding the subjectivity and errors of manual inspection. It can efficiently complete anomaly detection of large-scale point cloud data without human intervention. Furthermore, by calculating the maximum spatial distance between out-of-limit points and each surface of the baseline plate model, it further analyzes the distance relationship between these points and the model surface, ensuring the comprehensiveness and accuracy of contour detection. This ensures precise handling of various complex cargo shapes and flexible adaptation to diverse inspection standards, providing a revolutionary intelligent solution for the air cargo industry. It achieves digital traceability throughout the entire inspection process and improves inspection efficiency.

[0045] The container cargo loading contour detection device provided by the present invention is described below. The container cargo loading contour detection device described below can be referred to in correspondence with the container cargo loading contour detection method described above.

[0046] Figure 2 A schematic diagram of a container pallet cargo loading contour detection device is shown. The device includes:

[0047] Data acquisition module 21 acquires the spatial coordinate information of the surface of the goods packaging to obtain multiple test points;

[0048] The relationship determination module 22 determines the element topology relationship between each test point and the reference plate model of the container plate based on each test point and the previously acquired reference plate model. The reference plate model is obtained by isometric expansion based on the three-dimensional reference model corresponding to the container plate shape. The element topology relationship is used to characterize the spatial positional relationship between the corresponding test point and the basic geometric elements in the reference plate model.

[0049] The point determination module 23 determines the out-of-limit point when the point to be measured is located outside the spatial envelope of the reference plate model based on the unit topology relationship.

[0050] The contour detection module 24 determines the maximum spatial distance between each out-of-limit point and each surface of the reference plate model, based on each out-of-limit point, and obtains the contour detection result.

[0051] In this embodiment, the data acquisition module 21 includes: a data acquisition unit for acquiring point cloud, depth map and original spatial coordinates of containerized cargo packaging; a data fusion unit for interpolating and completing the depth map based on the original spatial coordinates and point cloud, and converting the interpolated and completed depth map into three-dimensional point cloud data to obtain fusion information; and a coordinate determination unit for obtaining spatial coordinate information of the cargo packaging surface based on the fusion information.

[0052] Furthermore, the data acquisition module 21 also includes: a denoising unit, which, before obtaining the spatial coordinate information of the cargo packaging surface based on the fusion information, establishes a point cloud topology network based on the fusion information, and uses curvature analysis to identify and remove discrete points with topological anomalies, thereby obtaining fusion information after removing discrete points and eliminating discrete interference points; and / or a target data extraction unit, which, based on the fusion information or the fusion information after removing discrete points, combines a spatial region extraction algorithm to extract the effective data region.

[0053] In an optional embodiment, the device further includes: a model invocation module, which, before determining the unit topology relationship between each test point and the reference plate type model based on each test point and in conjunction with the previously acquired reference plate type model of the container plate, invokes the corresponding three-dimensional reference model of the container plate from a preset parameter library according to the previously determined container plate type; wherein, the preset parameter library includes three-dimensional reference models of container plates of various plate types, and the three-dimensional reference model is constructed based on the corresponding plate type of the container plate; a correction module, which generates a spatial offset vector along the normal direction for each topological unit on the surface of the three-dimensional reference model, and performs smooth correction on the spatial offset vector to suppress the distortion caused by the displacement of the surface of the three-dimensional reference model; and a model generation module, which obtains the reference plate type model by using equidistant dilation based on the corrected spatial offset vector and the three-dimensional reference model.

[0054] Furthermore, the model generation module includes: a model generation unit that performs isometric expansion on the three-dimensional reference model based on the corrected spatial offset vector to generate a spatial envelope to adapt to goods packaging of different sizes; and an index creation unit that divides the internal space of the spatial envelope into tetrahedral elements and establishes a spatial location query index to obtain a reference plate model.

[0055] The relationship determination module 22 is used to: quickly locate the spatial envelope in which the test point may be located or the adjacent spatial envelope by using the spatial location query index, and use a preset geometric algorithm to determine whether the test point is located in the spatial envelope in which it may be located or the adjacent spatial envelope, so as to obtain the unit topology relationship between the corresponding test point and the reference plate model.

[0056] The contour detection module 24 includes: a clustering unit, which performs density-based clustering based on all out-of-limit points to separate multiple spatially independent out-of-limit region clusters; a feature fitting unit, which extracts and fits the principal plane of each out-of-limit region cluster to obtain multiple planar clusters with the same distance representation; a feature determination unit, which, for each planar cluster, traverses the points within the planar cluster to determine the maximum spatial distance and normal angle between each point within the planar cluster and each surface of the reference plate model; and a result acquisition unit, which obtains the contour detection result based on the maximum spatial distance and normal angle between each point within the planar cluster and each surface of the reference plate model.

[0057] Specifically, the result acquisition unit includes: a data selection subunit, which selects the minimum maximum spatial distance as the feature distance of the plane cluster based on the maximum spatial distance between each point in the plane cluster and each surface of the reference plate model; a mapping subunit, which determines the orientation to be mapped based on the normal angle between the point corresponding to the minimum maximum spatial distance and each surface of the reference plate model, combined with a preset mapping orientation rule, and maps the corresponding point to the surface of the reference plate model according to the orientation to be mapped, thereby obtaining the position orientation positioning information of the corresponding plane cluster; wherein, the preset mapping orientation rule is configured in advance based on the value range of different normal angles and their corresponding mapping orientations; and the result acquisition subunit obtains the contour detection result based on the feature distance of the plane cluster and the position orientation positioning information of the plane cluster.

[0058] In summary, this invention collects spatial coordinate information of the cargo packaging surface through a data acquisition module to comprehensively cover the packaging surface, avoiding analytical biases caused by missing data and improving the accuracy of subsequent inspections. Furthermore, a relationship determination module, combined with a baseline plate model obtained through equidistant expansion, quickly establishes a topological relationship with the points to be tested, avoiding complex geometric calculations and improving processing efficiency. The point judgment module then quickly identifies out-of-limit points, avoiding the subjectivity and errors of manual inspection. This eliminates the need for human intervention and efficiently completes anomaly detection of large-scale point cloud data. Finally, a contour detection module calculates the maximum spatial distance between out-of-limit points and each surface of the baseline plate model to further analyze the distance relationship between these points and the model surface, ensuring the comprehensiveness and accuracy of contour detection. This ensures precise handling of various complex cargo shapes and flexible adaptation to diverse inspection standards, providing a revolutionary intelligent solution for the air cargo industry. It achieves digital traceability throughout the entire inspection process and improves inspection efficiency.

[0059] Figure 3 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 3As shown, the electronic device may include: a processor 310, a communications interface 320, a memory 330, and a communications bus 340, wherein the processor 310, the communications interface 320, and the memory 330 communicate with each other through the communications bus 340. The processor 310 can call logical instructions in the memory 330 to execute a container cargo loading contour detection method. This method includes: acquiring spatial coordinate information of the cargo packaging surface to obtain multiple test points; determining the element topology relationship between each test point and the reference plate model based on each test point and a previously acquired reference plate model of the container; wherein the reference plate model is obtained by isometric expansion based on a three-dimensional reference model corresponding to the container plate shape, and the element topology relationship is used to characterize the spatial positional relationship between the corresponding test point and the basic geometric elements in the reference plate model; determining, based on the element topology relationship, that when a test point is located outside the spatial envelope of the reference plate model, an out-of-limit point is obtained; and determining, based on each out-of-limit point, the maximum spatial distance between the out-of-limit point and each surface of the reference plate model, to obtain the contour detection result.

[0060] Furthermore, the logical instructions in the aforementioned memory 330 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, essentially, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0061] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the container cargo loading contour detection method provided by the above methods. The method includes: acquiring spatial coordinate information of the cargo packaging surface to obtain multiple test points; determining the element topology relationship between each test point and the reference plate model based on each test point and a previously acquired reference plate model of the container; wherein the reference plate model is obtained by isometric expansion based on a three-dimensional reference model corresponding to the container plate shape, and the element topology relationship is used to characterize the spatial positional relationship between the corresponding test point and the basic geometric elements in the reference plate model; determining, based on the element topology relationship, when the test point is located outside the spatial envelope of the reference plate model, an out-of-limit point is obtained; determining, based on each out-of-limit point, the maximum spatial distance between the out-of-limit point and each surface of the reference plate model, and obtaining the contour detection result.

[0062] In another aspect, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon. When executed by a processor, the computer program implements the container cargo loading contour detection method provided by the above methods. The method includes: acquiring spatial coordinate information of the cargo packaging surface to obtain multiple test points; determining the element topology relationship between each test point and the reference plate model based on each test point and a previously acquired reference plate model of the container; wherein the reference plate model is obtained by isometric expansion based on a three-dimensional reference model corresponding to the container plate shape, and the element topology relationship is used to characterize the spatial positional relationship between the corresponding test point and the basic geometric elements in the reference plate model; determining, based on the element topology relationship, when the test point is located outside the spatial envelope of the reference plate model, an out-of-limit point is obtained; and determining, based on each out-of-limit point, the maximum spatial distance between the out-of-limit point and each surface of the reference plate model, to obtain the contour detection result.

[0063] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0064] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0065] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A palletized cargo load profile detection method, characterized by, The method comprises the following steps: Obtain spatial coordinate information of the surface of the cargo package to obtain a plurality of to-be-measured points; Determine the unit topological relationship between each to-be-measured point and the reference pallet model based on the to-be-measured points and the reference pallet model obtained in advance; wherein the reference pallet model is obtained by equidistantly inflating a three-dimensional reference model corresponding to the pallet type in advance, and the unit topological relationship is used to represent the spatial positional relationship between the corresponding to-be-measured point and the basic geometric unit in the reference pallet model; When the to-be-measured point is located outside the space envelope of the reference pallet model, determine the out-of-limit point based on the unit topological relationship; Determine the maximum spatial distance between each out-of-limit point and each surface of the reference pallet model to obtain the contour detection result; Determine the maximum spatial distance between each out-of-limit point and each surface of the reference pallet model to obtain the contour detection result, comprising: Perform density-based clustering based on all out-of-limit points to obtain out-of-limit region clusters; Extract and fit the main plane of each out-of-limit region cluster to obtain a plurality of plane clusters with the same distance representation; For each plane cluster, traverse the points in the plane cluster to determine the maximum spatial distance and the normal angle between each point in the plane cluster and each surface of the reference pallet model; Obtain the contour detection result based on the maximum spatial distance and the normal angle between each point in the plane cluster and each surface of the reference pallet model.

2. The pallet cargo loading profile detection method according to claim 1, characterized in that, Obtain the contour detection result based on the maximum spatial distance and the normal angle between each point in the plane cluster and each surface of the reference pallet model, comprising: Select the smallest maximum spatial distance as the characteristic distance of the plane cluster based on the maximum spatial distance between each point in the plane cluster and each surface of the reference pallet model; Determine the to-be-mapped orientation based on the normal angle between the point corresponding to the smallest maximum spatial distance and each surface of the reference pallet model, and map the corresponding point to the surface of the reference pallet model according to the to-be-mapped orientation to obtain the position and orientation positioning information of the corresponding plane cluster; wherein the preset mapping orientation rule is configured in advance according to the value range of different normal angles and the corresponding mapping orientation; Obtain the contour detection result based on the characteristic distance of the plane cluster and the position and orientation positioning information of the plane cluster.

3. The pallet cargo loading profile detection method according to claim 1, characterized in that, Before determining the unit topological relationship between each to-be-measured point and the reference pallet model based on the to-be-measured points and the reference pallet model obtained in advance, the method comprises the following steps: Call the three-dimensional reference model of the corresponding pallet in the preset parameter library based on the pallet type determined in advance; wherein the preset parameter library comprises the three-dimensional reference model of the pallet of multiple pallet types, and the three-dimensional reference model is constructed based on the pallet of the corresponding pallet type in advance; For each topological unit of the surface of the three-dimensional reference model, generate a spatial offset vector in the normal direction, and perform smoothing correction on the spatial offset vector; Obtain the reference pallet model by equidistant inflation based on the corrected spatial offset vector and the three-dimensional reference model.

4. The pallet cargo loading profile detection method according to claim 3, characterized in that, According to the modified spatial offset vector and the three-dimensional reference model, a reference plate model is obtained by isometric expansion, comprising: According to the modified spatial offset vector, the three-dimensional reference model is isometrically expanded to generate a spatial envelope; The internal space of the spatial envelope is divided into tetrahedral units, and a spatial position query index is established to obtain a reference plate model.

5. The pallet cargo loading profile detection method according to claim 1, wherein, Obtain the spatial coordinate information of the cargo packaging surface, comprising: Obtain the point cloud, depth map and original spatial coordinates of the container plate cargo packaging; According to the original spatial coordinates and the point cloud, the depth map is interpolated and completed, and the interpolated and completed depth map is converted into three-dimensional point cloud data to obtain fusion information; According to the fusion information, the spatial coordinate information of the cargo packaging surface is obtained.

6. The pallet cargo loading profile detection method according to claim 5, wherein, Before obtaining the spatial coordinate information of the cargo packaging surface according to the fusion information, comprising: According to the fusion information, a point cloud topology network is established, and curvature analysis is used to identify and remove topologically abnormal discrete points to obtain fusion information after removing discrete points; and / or, According to the fusion information or the fusion information after removing discrete points, combined with a spatial region extraction algorithm, an effective data region is extracted.

7. A palletized cargo load profile detection apparatus, characterized by, Comprising: A data acquisition module obtains the spatial coordinate information of the cargo packaging surface to obtain a plurality of test points; A relationship determination module determines the unit topology relationship between each test point and the reference plate model according to each test point and the previously obtained reference plate model of the container plate; wherein the reference plate model is obtained by isometric expansion of the three-dimensional reference model corresponding to the container plate type in advance, and the unit topology relationship is used to represent the spatial position relationship between the corresponding test point and the basic geometric unit in the reference plate model; A point position judgment module determines that the test point is located outside the spatial envelope of the reference plate model according to the unit topology relationship to obtain an out-of-limit point; A contour detection module determines the maximum distance between the out-of-limit point and each surface of the reference plate model according to each out-of-limit point to obtain a contour detection result; The contour detection module comprises: A clustering unit performs density-based clustering according to all out-of-limit points to obtain out-of-limit region clusters; A feature fitting unit extracts and fits the main plane of each out-of-limit region cluster to obtain a plurality of plane clusters with the same distance representation; A feature determination unit traverses the points in each plane cluster to determine the maximum distance and normal angle between each point in the plane cluster and each surface of the reference plate model; A result acquisition unit obtains the contour detection result according to the maximum distance and normal angle between each point in the plane cluster and each surface of the reference plate model.

8. An electronic device comprising a memory, a processor, and a computer program stored on the memory and running on the processor, characterized in that, The processor executes the computer program to implement the container plate cargo loading contour detection method of any one of claims 1 to 6. 9.A non-transitory computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the container plate cargo loading contour detection method of any one of claims 1 to 6.

Images