Vehicle-mounted object volume measurement method, device, and medium

Abstract

The embodiment of the application provides a vehicle-mounted object volume measurement method, device and medium, a plurality of vehicle point clouds are generated through a plurality of point cloud radars and a speed measurement module; features of the vehicle point clouds are extracted; point cloud coarse registration processing is performed according to distances of the plurality of point cloud radars and the plurality of vehicle point clouds, to obtain a transformation matrix range; a first transformation matrix for converting a source point cloud into a target point cloud is obtained according to the features of the vehicle point clouds; the first transformation matrix is optimized according to an inverted corner distance between the source point cloud and the target point cloud, to obtain a target transformation matrix located in the transformation matrix range; the vehicle point cloud is converted into a converted point cloud according to the target transformation matrix; and a vehicle-mounted object volume is obtained according to the converted point cloud; the method has a good point cloud registration effect, improves registration efficiency, and is beneficial to subsequent modeling and volume calculation of the vehicle-mounted object.

Description

TECHNICAL FIELD

[0001] Embodiments of the present application relate to, but are not limited to, the field of intelligent detection, and in particular to a vehicle-mounted object volume measurement method, device and medium. BACKGROUND

[0002] With the rapid development of the logistics industry and the acceleration of modernization, large transport vehicles are applied to cargo transportation, so it is necessary to calculate the volume of the vehicle-mounted objects of large transport vehicles. At present, the volume of the cargo transported by large transport vehicles is mainly measured by installing multiple 3D cameras for multi-angle shooting and laser radar to measure the length, width and height of the vehicle-mounted objects. However, due to the defects of the 3D camera itself, the measurement result has a large error. Moreover, the existing scheme for measuring the volume of the vehicle-mounted objects by laser radar has problems such as detection blind area and large error of the detection result. SUMMARY

[0003] The following is a summary of the subject matter described in detail herein. This summary is not intended to limit the scope of the claims.

[0004] Embodiments of the present application aim to at least solve one of the technical problems existing in the prior art, and provide a vehicle-mounted object volume measurement method, device and medium, which can have good point cloud registration effect and improve the registration efficiency.

[0005] In an embodiment of the first aspect of the application, a vehicle-mounted object volume measurement method comprises:

[0006] generating a plurality of vehicle point clouds by a plurality of point cloud radars and a speed measurement module;

[0007] extracting features of the vehicle point clouds;

[0008] performing point cloud coarse registration processing according to distances of the plurality of point cloud radars and the plurality of vehicle point clouds to obtain a transformation matrix range;

[0009] obtaining a first transformation matrix for converting a source point cloud into a target point cloud according to the features of the vehicle point clouds, the source point cloud being the vehicle point cloud, and the target point cloud being a point cloud in a target coordinate system;

[0010] optimizing the first transformation matrix according to an inverse corner distance between the source point cloud and the target point cloud to obtain a target transformation matrix located in the transformation matrix range;

[0011] converting the vehicle point cloud into a converted point cloud according to the target transformation matrix;

[0012] obtaining a vehicle-mounted object volume according to the converted point cloud.

[0013] Some embodiments of the first aspect of the application, wherein the plurality of vehicle point clouds are generated by a plurality of point cloud radars and a speed measurement module, comprise:

[0014] a plurality of radar frames are obtained from the plurality of point cloud radars;

[0015] distances between the radar frames at different positions are obtained from the speed measurement module;

[0016] a plurality of vehicle point clouds are generated from the radar frames and the distances.

[0017] Some embodiments of the first aspect of the application, after the plurality of vehicle point clouds are generated by a plurality of point cloud radars and a speed measurement module, comprise:

[0018] a predetermined number of data points are selected from the vehicle point clouds as sample points, and a sample model is fitted according to the sample points;

[0019] distance values of all data points to the sample model are calculated;

[0020] when the distance values meet a predetermined condition, the sample model is determined as a target model corresponding to the ground;

[0021] ground data points are determined from the vehicle point clouds according to the target model, and the ground data points are deleted.

[0022] Some embodiments of the first aspect of the application, after the plurality of vehicle point clouds are generated by a plurality of point cloud radars and a speed measurement module, comprise:

[0023] a data point of the vehicle point clouds is taken as a first marker, and data points with a distance less than a predetermined radius threshold from the first marker are taken as first candidate points;

[0024] a standard deviation of the first candidate points from the first marker is calculated;

[0025] first candidate points with a distance greater than a predetermined multiple of the standard deviation from the first marker are taken as noise points, and the noise points are deleted.

[0026] Some embodiments of the first aspect of the application, after the plurality of vehicle point clouds are generated by a plurality of point cloud radars and a speed measurement module, comprise:

[0027] data points with a distance less than a predetermined distance threshold from a seed point are selected as second candidate points from a field of the seed point, the seed point being any data point in the vehicle point clouds;

[0028] the second candidate points are added to a cluster, and the second candidate points are determined as new seed points;

[0029] When the number of data points in the cluster is equal to a preset number threshold, data points outside the cluster are taken as clutter data points of corresponding clutter, and the clutter data points are deleted.

[0030] In some embodiments of the first aspect of the application, the obtaining the first transformation matrix for transforming the source point cloud into the target point cloud according to the features of the vehicle point cloud comprises:

[0031] inputting the features of the vehicle point cloud into a network model comprising an encoder and a decoder to calculate a relative pose between the source point cloud and the target point cloud;

[0032] obtaining the first transformation matrix according to the relative pose.

[0033] In some embodiments of the first aspect of the application, the obtaining the target transformation matrix within the range of the transformation matrix according to the chamfer distance between the source point cloud and the target point cloud comprises:

[0034] obtaining the chamfer distance between the source point cloud and the target point cloud;

[0035] when the chamfer distance between the source point cloud and the target point cloud is less than a preset chamfer distance threshold, optimizing the first transformation matrix by minimizing the chamfer distance between the source point cloud and the target point cloud;

[0036] when the chamfer distance between the source point cloud and the target point cloud is greater than or equal to the preset chamfer distance threshold, and the first transformation matrix satisfies a registration end condition, obtaining the target transformation matrix within the range of the transformation matrix.

[0037] In some embodiments of the first aspect of the application, the obtaining the volume of the vehicle-mounted object according to the transformed point cloud comprises:

[0038] generating a triangular mesh connected to a plurality of transformed point cloud data points;

[0039] inserting a new transformed point cloud data point into the triangular mesh, determining all cavity triangles containing the new transformed point cloud data point from the triangular mesh, deleting the cavity triangles to form cavities, connecting the transformed point cloud data points corresponding to the cavities to the new transformed point cloud data point to form a new triangular mesh;

[0040] when all the transformed point cloud data points are traversed, obtaining a target triangular mesh;

[0041] obtaining the volume of the vehicle-mounted object according to the volume of the sub-triangles in the target triangular mesh.

[0042] In an embodiment of the second aspect of the present application, an electronic device comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the vehicle-mounted object volume measurement method as described above when executing the computer program.

[0043] In an embodiment of the third aspect of the present application, a computer-readable storage medium stores computer-executable instructions for executing the vehicle-mounted object volume measurement method as described above.

[0044] The above scheme has at least the following beneficial effects: multiple vehicle point clouds are generated by multiple point cloud radars and speed measurement modules; features of the vehicle point clouds are extracted; point cloud coarse registration processing is performed according to distances of the multiple point cloud radars and the multiple vehicle point clouds to obtain a transformation matrix range; a first transformation matrix for converting a source point cloud into a target point cloud is obtained according to features of the vehicle point clouds; the first transformation matrix is optimized according to an inverted corner distance between the source point cloud and the target point cloud to obtain a target transformation matrix located in the transformation matrix range; the vehicle point cloud is converted into a converted point cloud according to the target transformation matrix; and a vehicle-mounted object volume is obtained according to the converted point cloud. The scheme has good point cloud registration effect, improves registration efficiency, and is beneficial to subsequent modeling and volume calculation of the vehicle-mounted object. BRIEF DESCRIPTION OF DRAWINGS

[0045] The accompanying drawings are included to provide a further understanding of the technical scheme of the present application, and constitute a part of the specification, and are used together with the embodiments of the present application to explain the technical scheme of the present application, and do not constitute a limitation on the technical scheme of the present application.

[0046] Figure 1 is a step diagram of the vehicle-mounted object volume measurement method provided by the embodiments of the present application;

[0047] Figure 2 is a sub-step diagram of step S100;

[0048] Figure 3 is a step diagram of the ground separation step;

[0049] Figure 4 is a step diagram of the step of filtering and removing isolated noise points;

[0050] Figure 5 is a step diagram of the step of removing impurities;

[0051] Figure 6 is a sub-step diagram of step S500;

[0052] Figure 7 is a sub-step diagram of step S600;

[0053] Figure 8 is a sub-step diagram of step S700. Detailed Implementation

[0054] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0055] It should be noted that although functional modules are divided in the device schematic diagram and a logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the device or the order in the flowchart. The terms "first," "second," etc., in the specification, claims, or the aforementioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence.

[0056] The embodiments of this application will be further described below with reference to the accompanying drawings.

[0057] An embodiment of this application provides a method for measuring the volume of an object mounted on a vehicle.

[0058] Reference Figure 1 Methods for measuring the volume of objects on a vehicle, including but not limited to the following steps:

[0059] Step S100: Generate multiple vehicle point clouds using multiple point cloud radars and speed measurement modules;

[0060] Step S200: Preprocess the vehicle point cloud;

[0061] Step S300: Extract features from the vehicle point cloud;

[0062] Step S400: Perform coarse registration of point clouds based on the distances of multiple point cloud radars and multiple vehicle point clouds to obtain the range of the transformation matrix;

[0063] Step S500: Obtain a first transformation matrix for converting the source point cloud into the target point cloud based on the features of the vehicle point cloud.

[0064] Step S600: Optimize the first transformation matrix based on the chamfer distance between the source point cloud and the target point cloud to obtain the target transformation matrix within the range of the transformation matrix;

[0065] Step S700: Convert the vehicle point cloud into a transformed point cloud according to the target transformation matrix;

[0066] Step S800: Obtain the volume of the vehicle-mounted object based on the converted point cloud.

[0067] Reference Figure 2 For step S100, multiple vehicle point clouds are generated using multiple point cloud radars and speed measurement modules, including:

[0068] Step S110, obtaining multiple radar frames according to multiple point cloud radars;

[0069] Step S120, obtaining the distance between radar frames at different positions according to the speed measurement module;

[0070] Step S130, generating multiple vehicle point clouds according to the radar frames and the distance.

[0071] For step S110, the vehicle is controlled to pass through the collection area at low speed, and point cloud radars are arranged at two different positions. Specifically, the point cloud radars are laser radars. Radar frames of the same vehicle at different times are obtained by the point cloud radars at the two different positions.

[0072] For step S120, the distance of the vehicle forward displacement between two radar frames is obtained by the speed measurement module.

[0073] For step S130, multiple vehicle point clouds are generated according to the radar frames and the distance. Specifically, empty vehicle point clouds of empty vehicles and full vehicle point clouds of full vehicles are generated to improve the accuracy of volume calculation.

[0074] By laser radars cooperating with the speed measurement module, real-time speed measurement ensures the displacement between two frames, and the authenticity of the length-width-height ratio of the point cloud is ensured.

[0075] For step S200, the vehicle point cloud is preprocessed, which includes the steps of separating the ground, filtering and removing isolated noise points, and removing debris.

[0076] Reference Figure 3 The ground separation step includes but is not limited to the following steps:

[0077] Step S211, selecting a preset number of data points from the vehicle point cloud as sample points, and fitting a sample model according to the sample points;

[0078] Step S212, calculating the distance value of all data points to the sample model;

[0079] Step S213, when the distance value meets the preset condition, determining the sample model as a target model corresponding to the ground;

[0080] Step S214, determining the ground data points from the vehicle point cloud according to the target model, and deleting the ground data points.

[0081] For step S211, two or other preset number of data points are randomly selected from the vehicle point cloud as sample points, and a sample model is fitted according to the sample points. The sample model can be a straight line or other graphics.

[0082] For step S212, the distance value of all data points to the straight line is calculated.

[0083] For step S213, it can be understood that if the distance value is less than the threshold s, it is considered that the data point is on the sample model, and if the distance value is greater than the threshold s, it is considered that the data point is not on the sample model.

[0084] When the distance value meets the preset condition, for example, the distance value meets the loss function, or the number of data points on the sample model is greater than the preset number threshold, the sample model is determined as the target model corresponding to the ground. If the distance value does not meet the preset condition, steps S211 and S212 are repeated.

[0085] For step S214, the ground data points are determined from the vehicle point cloud according to the target model, and the ground data points are deleted, and the data points related to the vehicle are retained as much as possible.

[0086] Referring to Figure 4 , the filtering step of removing isolated noise points includes but is not limited to the following steps:

[0087] The following steps are performed for each data point of the vehicle point cloud:

[0088] Step S221, taking a data point of the vehicle point cloud as a first marker, and taking data points with a distance less than a preset radius threshold from the first marker as first candidate points;

[0089] Step S222, calculating the standard deviation of the first candidate points and the first marker;

[0090] Step S223, taking the first candidate points with a distance greater than a preset multiple of the standard deviation from the first marker as noise points, and deleting the noise points.

[0091] Referring to Figure 5 , the step of removing debris includes but is not limited to the following steps:

[0092] Step S231, selecting data points with a distance less than a preset distance threshold from a seed point as second candidate points from the field of the seed point, the seed point being any data point in the vehicle point cloud;

[0093] Step S232, adding the second candidate points to the cluster, and determining the second candidate points as new seed points;

[0094] Steps S231 and S232 are repeatedly executed.

[0095] Step S233, when the number of data points in the cluster is equal to a preset number threshold, taking data points outside the cluster as debris data points corresponding to the debris, and deleting the debris data points.

[0096] The data points of the pre-processed vehicle point cloud are used for point cloud registration processing. Point cloud registration is one of the important steps in point cloud processing, which aligns point cloud data of different positions, orientations or sizes to the same coordinate system, making subsequent point cloud segmentation, target recognition and other processing more accurate and reliable. The point cloud data after registration can establish a more accurate model.

[0097] For step S300, the pre-processed vehicle point cloud is subjected to farthest distance sampling and geometric sampling, then the sample data obtained by farthest distance sampling and the sample data obtained by geometric sampling are combined, and farthest distance sampling is performed again to reduce the sample density and effectively improve the registration efficiency. The sample data obtained by sampling can retain the main features of the point cloud, and also retain the contour features of the geometric sampling.

[0098] The features of each vehicle point cloud are extracted; the features of the vehicle point cloud are usually the coordinates, color, normal vector and other attributes of the data points.

[0099] For step S400, the nearest neighbor search algorithm or other algorithms are used to find the neighboring data points of each data point in each vehicle point cloud.

[0100] Some initial registration methods (such as ICP or feature-based methods) are used to perform initial registration on the vehicle point cloud, which will generate a rough transformation matrix. According to the distance of the multiple point cloud radars and the multiple vehicle point clouds, point cloud coarse registration processing is performed to obtain a transformation matrix range. The transformation matrix can convert the coordinate system of one point cloud to the coordinate system of another point cloud.

[0101] According to the positional relationship of the two laser radars, the registration error can be controlled within a certain range in advance, which provides a guarantee for subsequent fine registration and greatly reduces the situation that two point clouds fall into local optimal solution due to large offset angle or large displacement. At the same time, it also reduces the difficulty of subsequent registration and improves the registration efficiency.

[0102] Reference Figure 6 For step S500, a first transformation matrix for converting the source point cloud to the target point cloud is obtained according to the features of the vehicle point cloud, including but not limited to the following steps:

[0103] Step S510, input the features of the vehicle point cloud into a network model including an encoder and a decoder to calculate the relative pose between the source point cloud and the target point cloud;

[0104] Step S520, obtaining the first transformation matrix according to the relative pose.

[0105] Wherein, the source point cloud is a vehicle point cloud, and the target point cloud is a point cloud in a target coordinate system.

[0106] Specifically, the network model is RPMnet, which is a neural network model composed of an encoder and a decoder. The encoder takes the features of the vehicle point cloud as input and inputs an intermediate representation. The decoder takes the intermediate representation as input, calculates the relative pose between the source point cloud and the target point cloud, and outputs a first transformation matrix according to the relative pose.

[0107] Referring to Figure 7 For step S600, the first transformation matrix is optimized according to the chamfer distance between the source point cloud and the target point cloud to obtain a target transformation matrix within the range of the transformation matrix, including but not limited to the following steps:

[0108] Step S610, obtaining the chamfer distance between the source point cloud and the target point cloud;

[0109] Step S620, when the chamfer distance between the source point cloud and the target point cloud is less than a preset chamfer distance threshold, the first transformation matrix is optimized by minimizing the chamfer distance between the source point cloud and the target point cloud;

[0110] Step S630, when the chamfer distance between the source point cloud and the target point cloud is greater than or equal to the preset chamfer distance threshold, and the first transformation matrix satisfies the registration end condition, a target transformation matrix within the range of the transformation matrix is obtained.

[0111] For step S620, the parameters of the first transformation matrix can be optimized by minimizing the chamfer distance between the source point cloud and the target point cloud through an optimization algorithm such as the Levenberg-Marquardt algorithm.

[0112] By iterating steps S610 and S620 multiple times, when the chamfer distance between the source point cloud and the target point cloud is greater than or equal to the preset chamfer distance threshold, and the first transformation matrix satisfies the registration end condition, a target transformation matrix is obtained.

[0113] Referring to Figure 8 For step S700, the vehicle object volume is obtained according to the transformed point cloud, including but not limited to the following steps:

[0114] Step S710, generating a triangular mesh connected to a plurality of transformed point cloud data points;

[0115] Step S720, inserting a new transformed point cloud data point into the triangular mesh, determining all circumcircle triangles containing the new transformed point cloud data point from the triangular mesh, deleting the circumcircle triangles to form a cavity, and connecting the transformed point cloud data points corresponding to the cavity to the new transformed point cloud data point to form a new triangular mesh;

[0116] Step S730, when all transformed point cloud data points are traversed, a target triangular mesh is obtained;

[0117] Step S740, obtaining the volume of the vehicle-mounted object according to the volume of the sub-triangle in the target triangular mesh.

[0118] For step S710, a plurality of converted point cloud data points are randomly selected, lines are connected between the converted point cloud data points, and a triangular mesh is generated.

[0119] For step S720, a new converted point cloud data point is inserted into the triangular mesh, all cavity triangles containing the new converted point cloud data point are determined from the triangular mesh or all cavity triangles having a common side with the triangle containing the new converted point cloud data point are found, the cavity triangles are deleted to form cavities, the converted point cloud data points corresponding to the cavities are connected with the new converted point cloud data point, and a new triangular mesh is formed.

[0120] For step S730, step S720 is repeated until all converted point cloud data points are inserted into the triangular mesh, and a target triangular mesh is obtained.

[0121] For step S740, the volumes of all sub-triangles in the target triangular mesh are calculated, the sum of the volumes of all sub-triangles is calculated, and the sum of the volumes of all sub-triangles is taken as the volume of the vehicle-mounted object.

[0122] Embodiments of the present application provide an electronic device. The electronic device comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the vehicle-mounted object volume measurement method as described above when executing the computer program.

[0123] The electronic device can be any smart terminal including a tablet computer, a vehicle-mounted computer, etc.

[0124] Generally, for the hardware structure of the electronic device, the processor can be implemented in the form of a general-purpose CPU (Central Processing Unit), a microprocessor, an ASIC (Application Specific Integrated Circuit), or one or more integrated circuits, etc., for executing related programs to implement the technical solutions provided by the embodiments of the present application.

[0125] The memory can be implemented in the form of a ROM (ReadOnly Memory), a static storage device, a dynamic storage device, or a RAM (Random Access Memory), etc. The memory can store an operating system and other application programs, and when the technical solutions provided by the embodiments of the present application are implemented by software or firmware, the related program codes are saved in the memory and are called and executed by the processor to implement the method of the embodiments of the present application.

[0126] The input / output interface is configured to realize information input and output.

[0127] The communication interface is configured to realize communication interaction between the device and other devices. The communication can be realized through a wired manner (for example, a USB, a network cable, etc.) or a wireless manner (for example, a mobile network, WIFI, Bluetooth, etc.).

[0128] The bus transmits information between various components (for example, the processor, the memory, the input / output interface and the communication interface) of the device. The processor, the memory, the input / output interface and the communication interface are connected to each other through the bus to realize communication connection within the device.

[0129] The embodiment of the present application provides a computer readable storage medium. The computer readable storage medium stores computer executable instructions. The computer executable instructions are used to execute the vehicle-mounted object volume measurement method.

[0130] It should be appreciated that the method steps in the embodiments of the present application can be realized or implemented by computer hardware, a combination of hardware and software, or through computer instructions stored in a non-transitory computer readable storage medium. The method can use standard programming techniques. Each program can be implemented in a high-level procedural or object-oriented programming language to communicate with a computer system. However, if necessary, the program can be implemented in an assembly or machine language. In any case, the language can be a compiled or interpreted language. In addition, the program can run on a programmed special integrated circuit for this purpose.

[0131] In addition, the operations of the processes described herein can be performed in any suitable order unless otherwise indicated herein or otherwise clearly contradicted by context. The processes described herein (or variations and / or combinations thereof) can be performed under the control of one or more computer systems configured with executable instructions (e.g., executable instructions, one or more computer programs or one or more applications), and can be implemented as code (e.g., executable instructions, one or more computer programs or one or more applications) executing collectively on one or more processors, by hardware or combinations thereof. The computer programs include a plurality of instructions executable by one or more processors.

[0132] Further, the methods can be implemented in any type of computing platform operably connected to a suitable computing platform, including but not limited to a personal computer, a smart phone, a mainframe, a workstation, a network or distributed computing environment, a stand-alone or integrated computer platform, or in communication with a charged particle tool or other imaging device, and the like. Aspects of the present application can be implemented in machine readable code stored on a non-transitory storage medium or device, whether removable or integrated into a computing platform, such as a hard disk, an optical read and / or write storage medium, RAM, ROM, and the like, such that it can be read by a programmable computer to configure and operate the computer to perform the processes described herein when the storage medium or device is read by the computer. In addition, the machine readable code, or portions thereof, can be transmitted over wired or wireless networks. The present application described herein includes these and other different types of non-transitory computer readable storage media when including instructions or programs that implement the steps described above in connection with microprocessors or other data processors. The present application also includes the computer itself when programmed according to the methods and techniques described in connection with the present application.

[0133] The computer program can be applied to input data to perform the functions described herein to transform the input data to generate output data that is stored to non-volatile memory. The output information can also be applied to one or more output devices such as a display. In a preferred embodiment of the present application, the transformed data represents a physical and tangible object, including a specific visual depiction of the physical and tangible object produced on a display.

[0134] While the embodiments of the application have been shown and described, it is to be understood that the application is not limited to the details of the embodiments described, since the generic principles of the application can be employed to other embodiments without departing from the spirit and scope of the application as defined by the appended claims and their equivalents.

[0135] The above is a specific description of the preferred embodiments of the application, but the application is not limited to the embodiments, and those skilled in the art can make various equivalent modifications or replacements without departing from the spirit of the application. These equivalent modifications or replacements are all included in the scope defined by the embodiments.

Claims

1. A method of measuring the volume of an object in a vehicle, characterized by, include: Multiple vehicle point clouds are generated using multiple point cloud radars and speed measurement modules. Extract features from vehicle point clouds; Based on the distances of multiple point cloud radars and multiple vehicle point clouds, coarse registration of the point clouds is performed to obtain the range of the transformation matrix; A first transformation matrix is ​​obtained based on the characteristics of the vehicle point cloud to convert the source point cloud into a target point cloud, wherein the source point cloud is the vehicle point cloud and the target point cloud is the point cloud in the target coordinate system; The first transformation matrix is ​​optimized based on the chamfer distance between the source point cloud and the target point cloud to obtain a target transformation matrix within the range of the transformation matrix; The vehicle point cloud is converted into a transformed point cloud based on the target transformation matrix. The volume of the vehicle-mounted object is obtained from the converted point cloud. The step of obtaining the volume of the vehicle-mounted object based on the converted point cloud includes: Generate a triangular mesh connecting several transformed point cloud data points; Insert new converted point cloud data points into the triangular mesh, determine all cavity triangles whose circumcircles contain the new converted point cloud data points from the triangular mesh, delete the cavity triangles to form cavities, and connect the converted point cloud data points corresponding to the cavities with the new converted point cloud data points to form a new triangular mesh. When all the transformed point cloud data points are traversed, the target triangular mesh is obtained; The volume of the vehicle-mounted object is obtained based on the volume of the sub-triangles in the target triangular mesh.

2. The on-vehicle object volume measurement method according to claim 1, characterized by, The process of generating multiple vehicle point clouds using multiple point cloud radars and speed measurement modules includes: Multiple radar frames are acquired based on the multiple point cloud radars; The distance between radar frames at different locations is obtained based on the speed measurement module; Multiple vehicle point clouds are generated based on the radar frame and the distance.

3. The on-vehicle object volume measurement method according to claim 1, characterized by, After generating multiple vehicle point clouds using multiple point cloud radars and speed measurement modules, the process includes: A preset number of data points are selected from the vehicle point cloud as sample points, and a sample model is obtained by fitting the sample points. Calculate the distance from all data points to the sample model; When the distance value meets the preset conditions, the sample model is determined as the target model of the corresponding ground. Based on the target model, ground data points are determined from the vehicle point cloud, and then the ground data points are deleted.

4. The on-vehicle object volume measurement method according to claim 1, characterized by, After generating multiple vehicle point clouds using multiple point cloud radars and speed measurement modules, the process includes: Take one data point of the vehicle point cloud as the first marker, and take data points whose distance from the first marker is less than a preset radius threshold as the first candidate points; Calculate the standard deviation between the first candidate point and the first punctuation point; First candidate points whose distance from the first punctuation mark is greater than a preset multiple of the standard deviation of the first punctuation mark are identified as noise points and deleted.

5. The on-vehicle object volume measurement method according to claim 1, characterized by, After generating multiple vehicle point clouds using multiple point cloud radars and speed measurement modules, the process includes: From the neighborhood of the seed point, select data points whose distance from the seed point is less than a preset distance threshold as second candidate points, where the seed point is any data point in the vehicle point cloud; Add the second candidate point to the cluster and determine the second candidate point as the new seed point; When the number of data points in the cluster is equal to a preset number threshold, data points outside the cluster are taken as clutter data points of corresponding clutters, and the clutter data points are deleted.

6. The on-vehicle object volume measurement method according to claim 1, characterized by, The first transformation matrix for converting the source point cloud into the target point cloud is obtained according to the features of the vehicle point cloud, and the method comprises the steps of: inputting the features of the vehicle point cloud into a network model comprising an encoder and a decoder to calculate the relative pose between the source point cloud and the target point cloud; obtaining the first transformation matrix according to the relative pose.

7. The on-vehicle object volume measurement method according to claim 1, characterized by, The target transformation matrix within the range of the transformation matrix is obtained by optimizing the first transformation matrix according to the chamfer distance between the source point cloud and the target point cloud, and the method comprises the steps of: obtaining the chamfer distance between the source point cloud and the target point cloud; when the chamfer distance between the source point cloud and the target point cloud is less than a preset chamfer distance threshold, the first transformation matrix is optimized by minimizing the chamfer distance between the source point cloud and the target point cloud; when the chamfer distance between the source point cloud and the target point cloud is greater than or equal to the preset chamfer distance threshold, and the first transformation matrix satisfies the registration end condition, the target transformation matrix within the range of the transformation matrix is obtained.

8. An electronic device comprising: A memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the vehicle-mounted object volume measurement method according to any one of claims 1 to 7 when executing the computer program.

9. A computer-readable storage medium, characterized in that, A computer executable instruction is stored, and the computer executable instruction is used to execute the vehicle-mounted object volume measurement method according to any one of claims 1 to 7.

Images