Scrap steel scattered material identification method and device, electronic equipment and storage medium

Abstract

The invention provides a waste steel scattered material identification method and device, electronic equipment and a storage medium, and the method comprises the steps: obtaining scene point cloud samples, carrying out the feature marking, and constructing a training data set specially for waste steel scattered materials, therefore, the scattered scrap steel in the target area can be identified more accurately, the identification precision and the identification efficiency of the scattered scrap steel are improved, the whole identification process is highly automatic, manual intervention is reduced from data acquisition and calibration to model identification, the working efficiency is improved, and the identification accuracy and the identification efficiency of the scattered scrap steel are improved. According to the method, the position information of all the scattered materials can be accurately extracted, the position information comprises the small scattered materials which are difficult to cut away from the ground, the height information of the scattered materials can be accurately measured, the comprehensiveness and accuracy of scattered material stacking operation are ensured, the operation safety is improved, and potential risks caused by missing or misjudgment of the scattered materials are avoided.

Description

Technical Field

[0001] The present invention relates to the field of steel intelligent sensing technology, and in particular to a method, device, electronic equipment and storage medium for identifying scattered scrap steel. Background Art

[0002] As a high-quality renewable resource, scrap steel is one of the important raw materials in the steel smelting industry. Compared with the production of steel using iron ore as raw material, the efficiency of steelmaking from scrap steel is significantly improved and harmful emissions are significantly reduced, so it has high ecological and social value. Intelligent warehouse management technology for scrap steel includes key technical links such as unmanned vehicle loading and unloading, scattered material stacking, and warehouse inventory. Among them, scattered material stacking refers to the unmanned vehicle cleaning up the scrap steel materials scattered in the parking area during the loading and unloading process after the scrap steel truck leaves the parking space. The unmanned operation of stacking scattered materials is highly dependent on visual perception to obtain scattered material information.

[0003] At present, the existing scattered material recognition methods have the problem of insufficient recognition information or low accuracy. For example, the existing scattered material recognition methods based on two-dimensional images can only obtain the location information of scattered materials, but cannot obtain their height information. The existing scattered material recognition methods based on three-dimensional laser point clouds using the PCL classic algorithm can obtain the height information of scattered materials, but it is difficult to accurately extract all scattered materials, especially when the scattered materials are small and difficult to separate from the ground, and it is impossible to accurately obtain the location and height information of all scattered materials at the same time. Summary of the invention

[0004] The present application provides a method, device, electronic device and storage medium for identifying scattered scrap steel to solve the above-mentioned technical problems.

[0005] A method for identifying scattered scrap provided in an embodiment of the present application comprises: acquiring scene point cloud samples, and performing feature annotation on the scene point cloud samples to obtain a training data set; performing training based on a selected neural network through the training data set to obtain a point cloud segmentation model for scattered scrap; collecting three-dimensional point cloud data of a target area, and calibrating the three-dimensional point cloud data to obtain three-dimensional point cloud data to be detected, wherein the target area includes a designated parking space area and a margin area around the parking space, and the margin area includes a space within a preset range beyond the boundary of the designated parking space area; inputting the three-dimensional point cloud data to be detected into the point cloud segmentation model for scattered scrap to obtain a scattered material point cloud of the target area, and computing the scattered material grid information of the target area through the scattered material point cloud to complete the identification of scattered scrap.

[0006] In one embodiment of the present application, scene point cloud samples are obtained, and feature annotation is performed on the scene point cloud samples to obtain a training data set, including: collecting a scattered material point cloud sample set, the scattered material point cloud sample set includes a variety of scattered material forms with different distributions; irregularly marking along the edge of each scattered material pile, assigning a first label value to the scattered material, and assigning a second label value to non-scattered material points to obtain a training data set.

[0007] In one embodiment of the present application, three-dimensional point cloud data of the target area is collected, and the point cloud is calibrated to obtain three-dimensional point cloud data to be detected, including: performing three-dimensional laser scanning on the target area to obtain an initial point cloud of the target area; calibrating and calibrating the initial point cloud data through scanner calibration parameters to obtain calibrated point cloud data; the scanner calibration parameters are used to correct scanner installation errors; and through-filtering the calibrated point cloud data according to a parking area threshold to filter out point clouds in non-parking areas to obtain three-dimensional point cloud data to be detected.

[0008] In one embodiment of the present application, the three-dimensional point cloud data to be detected is input into a scattered scrap material point cloud segmentation model to obtain a scattered material point cloud of the target area, including: using the scattered material segmentation model to deduce the three-dimensional point cloud data to be detected to obtain a deduction result, wherein the deduction result includes a label value for each point in the three-dimensional point cloud data to be detected in a specified parking space area; traversing each point and its label value in the deduction result of the scattered material segmentation model, and synthesizing the scattered material point cloud with a label value as the point.

[0009] In one embodiment of the present application, the scattered material grid information of the target area is calculated by the scattered material point cloud, and also includes: gridding the scattered material point cloud in the target area according to preset grid specification parameters to obtain multiple grid areas, and the grid area includes grid parameters; through-filtering the scattered material point cloud in the target area through the grid area parameters to obtain a grid scattered point cloud; traversing the point cloud height values ​​of each grid scattered point cloud, and recording the maximum height value point in each grid scattered point cloud as the grid maximum height value point; determining the grid area parameters and the grid maximum height value point as the scattered material grid information of the target area.

[0010] In one embodiment of the present application, the scattered material point cloud in the target area is gridded according to preset grid specification parameters to obtain a plurality of grid areas, and the grid areas include grid area parameters, including: if the target area is a parking space area with a rectangular shape, a spatial rectangular coordinate system is established, and the vertex angles of the parking area are marked to obtain the vertex angle coordinates of the parking area; the marked parking space area is gridded according to the preset grid specification parameters to obtain a plurality of grid areas, the diagonal coordinates of the grid area are calculated according to the vertex angle coordinates and the preset grid specification parameters, and the grid area parameters of each grid area are marked according to the diagonal coordinates.

[0011] In one embodiment of the present application, calibrating the scanner calibration parameters includes: performing an initial scan on the target area to obtain three-dimensional point cloud data; placing a point cloud calibration marker in the target area, performing a secondary scan on the target area to obtain a scene point cloud containing the marker; and calculating the scanner calibration parameters through the three-dimensional point cloud data and the scene point cloud containing the marker.

[0012] The embodiment of the present application provides a device for identifying scattered scrap steel, and the device includes: an acquisition module, which is used to acquire scene point cloud samples, perform feature annotation on the scene point cloud samples, and obtain a training data set; a training module, which is used to perform training based on a selected neural network through the training data set to obtain a point cloud segmentation model for scattered scrap steel; a data calibration module, which is used to collect three-dimensional point cloud data of a target area, and calibrate the three-dimensional point cloud data to obtain three-dimensional point cloud data to be detected, wherein the target area includes a designated parking space area and a margin area around the parking space, and the margin area includes a space within a preset range beyond the boundary of the designated parking space area; an identification module, which is used to input the three-dimensional point cloud data to be detected into the point cloud segmentation model for scattered scrap steel, obtain a scattered material point cloud of the target area, and calculate the scattered material grid information of the target area through the scattered material point cloud to complete the identification of scattered scrap steel.

[0013] An embodiment of the present application provides an electronic device, including a processor, a memory and a communication bus; the communication bus is used to connect the processor and the memory; the processor is used to execute a computer program stored in the memory to implement the above-mentioned scrap scattered material identification method.

[0014] An embodiment of the present application provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program is used to enable a computer to execute the above-mentioned method for identifying scattered scrap steel.

[0015] The beneficial effects of the present application are as follows: by acquiring scene point cloud samples and performing feature annotation, a training data set specifically for scattered scrap steel is constructed; the scattered scrap steel point cloud segmentation model trained based on this data set can more accurately identify scattered scrap steel in the target area, thereby improving the recognition accuracy and efficiency of scattered scrap steel; the entire recognition process is highly automated, from data collection, calibration to model recognition, which reduces manual intervention and improves work efficiency; and the present application can not only accurately extract the location information of all scattered materials, including those that are smaller and difficult to separate from the ground, but also accurately measure their height information, ensuring the comprehensiveness and accuracy of the scattered material stacking operation, but also greatly improving the safety of the operation, avoiding potential risks caused by missing or misjudging scattered materials; in addition, the present application can also change the path planning and operation strategies that help optimize unmanned vehicles, further reduce energy consumption and improve overall operating efficiency.

[0016] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The drawings herein are incorporated into the specification and constitute a part of the specification, showing embodiments consistent with the present application, and together with the specification, are used to explain the principles of the present application. Obviously, the drawings described below are only some embodiments of the present application, and for those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative work. In the drawings:

[0018] Figure 1 is a flow chart of a method for identifying scattered scrap steel shown in an exemplary embodiment of the present application;

[0019] Figure 2 is a schematic diagram of scanner installation shown in an exemplary embodiment of the present application;

[0020] Figure 3 is a schematic diagram of a parking space area grid processing shown in an exemplary embodiment of the present application;

[0021] Figure 4 is a block diagram of a device for identifying scattered scrap steel shown in an exemplary embodiment of the present application;

[0022] Figure 5 It is a structural schematic diagram of an electronic device shown in an exemplary embodiment of the present application. DETAILED DESCRIPTION

[0023] The following will describe the implementation methods of the present application with reference to the accompanying drawings and preferred embodiments. Those skilled in the art can easily understand other advantages and effects of the present application from the contents disclosed in this specification. The present application can also be implemented or applied through other different specific implementation methods, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present application. It should be understood that the preferred embodiments are only for illustrating the present application, not for limiting the scope of protection of the present application.

[0024] It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present application, and thus the drawings only show components related to the present application rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed at will, and the component layout may also be more complicated.

[0025] In the following description, a large number of details are discussed to provide a more thorough explanation of the embodiments of the present application. However, it is obvious to those skilled in the art that the embodiments of the present application can be implemented without these specific details. In other embodiments, well-known structures and devices are shown in the form of block diagrams rather than in detail to avoid making the embodiments of the present application difficult to understand.

[0026] The embodiments of the present application respectively provide a method for identifying scattered scrap steel, a device for identifying scattered scrap steel, and an electronic device, which will be described in detail below.

[0027] See also Figure 1 , Figure 1 FIG. 1 is a flow chart of a method for identifying scattered scrap steel according to an embodiment of the present application. Figure 1 As shown, the method at least includes steps S110 to S140, which are described in detail as follows:

[0028] Step S110, obtaining scene point cloud samples, and performing feature annotation on the scene point cloud samples to obtain a training data set.

[0029] In one embodiment of the present application, Figure 2 FIG. 1 is a schematic diagram of a scanner installation shown in an exemplary embodiment of the present application. Figure 2 As shown, a scanner is installed in the scrap steel storage area, and the scanner is used to scan the designated parking area in the designated area to obtain a scene point cloud sample.

[0030] Step S120, training is performed based on the selected neural network using the training data set to obtain a point cloud segmentation model for scattered scrap steel.

[0031] In one embodiment of the present application, the selected neural network may be a Pointnet++ network, a model is built by using the Pointnet++ network, and training is performed using a training data set to obtain a point cloud segmentation model for scattered scrap steel.

[0032] Step S130, collect three-dimensional point cloud data of the target area, and calibrate the three-dimensional point cloud data to obtain the three-dimensional point cloud data to be detected. The target area includes the designated parking area and the margin area around the parking space. The margin area includes the space within a preset range beyond the boundary of the designated parking area.

[0033] Step S140, input the three-dimensional point cloud data to be detected into the scattered scrap material point cloud segmentation model to obtain the scattered material point cloud of the target area, and calculate the scattered material grid information of the target area through the scattered material point cloud to complete the scattered scrap material identification.

[0034] exist Figure 1 In the technical solution shown, a training data set specifically for scattered scrap steel is constructed by acquiring scene point cloud samples and performing feature annotation. The scattered scrap steel point cloud segmentation model trained based on this data set can more accurately identify scattered scrap steel in the target area, thereby improving the recognition accuracy and efficiency of scattered scrap steel. The entire recognition process is highly automated, from data collection, calibration to model recognition, which reduces manual intervention and improves work efficiency. In addition, the present application can not only accurately extract the location information of all scattered materials, including those that are smaller and difficult to separate from the ground, but also accurately measure their height information, ensuring the comprehensiveness and accuracy of the scattered material stacking operation, but also greatly improving the safety of the operation and avoiding potential risks caused by missing or misjudging scattered materials. In addition, the present application can also change the path planning and operation strategy that helps optimize unmanned vehicles, further reduce energy consumption and improve overall operation efficiency.

[0035] In one embodiment of the present application, scene point cloud samples are obtained, and feature annotation is performed on the scene point cloud samples to obtain a training data set, including: collecting a scattered material point cloud sample set, the scattered material point cloud sample set includes a variety of scattered material forms with different distributions; irregularly marking along the edge of each scattered material pile, assigning a first label value to the scattered material, and assigning a second label value to non-scattered material points to obtain a training data set.

[0036] As an example, point cloud sample data of scrap scattered material scenes in a large number of parking spaces are collected in a designated parking area. The point cloud sample data of the scrap scattered material scene contains a variety of scattered material forms with different distributions. The sample annotation is performed using labelme software. The annotation method is to irregularly mark along the edge of each scattered material pile, and assign a first label value to the scattered material, for example, the first label value is "1", and the non-scattered material points are assigned a second label value, for example, the second label value is "2", so as to obtain a training data set.

[0037] In one embodiment of the present application, three-dimensional point cloud data of a target area is collected, and the point cloud is calibrated to obtain three-dimensional point cloud data to be detected, including: performing three-dimensional laser scanning on the target area to obtain an initial point cloud of the target area; calibrating and calibrating the initial point cloud data through scanner calibration parameters to obtain calibrated point cloud data; the scanner calibration parameters are used to correct scanner installation errors; and the calibrated point cloud data is directly filtered according to a parking area threshold to filter out point clouds in non-parking areas to obtain three-dimensional point cloud data to be detected.

[0038] In one embodiment of the present application, the three-dimensional point cloud data to be detected is input into a point cloud segmentation model for scattered scrap steel to obtain a scattered material point cloud in the target area, including: using the scattered material segmentation model to deduce the three-dimensional point cloud data to be detected to obtain a deduction result, the deduction result including the label value of each point in the three-dimensional point cloud data to be detected in the specified parking space area; traversing each point and its label value in the deduction result of the scattered material segmentation model, and synthesizing the scattered material point cloud with a label value as the point.

[0039] As an example, the three-dimensional point cloud data to be detected is input into the scattered material point cloud segmentation model of scrap steel to obtain the scattered material point cloud of the target area. That is, the scattered material segmentation model is used to deduce the three-dimensional point cloud data to be detected, and the index and label value of each point in the output result of the scattered material segmentation model are traversed, and all points with label value "1" are extracted, and these points are synthesized into a scattered material point cloud.

[0040] In one embodiment of the present application, the scattered material grid information of the target area is calculated by the scattered material point cloud, which also includes: gridding the scattered material point cloud in the target area according to preset grid specification parameters to obtain multiple grid areas, and the grid areas include grid area parameters; through-filtering the scattered material point cloud in the target area through the grid area parameters to obtain a grid scattered point cloud; traversing the point cloud height values ​​of each grid scattered point cloud, and recording the maximum height value point in each grid scattered point cloud as the grid maximum height value point; determining the grid area parameters and the grid maximum height value point as the scattered material grid information of the target area.

[0041] In one embodiment of the present application, the scattered material point cloud in the target area is gridded according to preset grid specification parameters to obtain multiple grid areas, and the grid areas include grid area parameters, including: if the target area is a parking space area with a rectangular shape, a spatial rectangular coordinate system is established, and the vertex angles of the parking area are marked to obtain the vertex angle coordinates of the parking area; the marked parking space area is gridded according to the preset grid specification parameters to obtain multiple grid areas, the diagonal coordinates of the grid area are calculated according to the vertex angle coordinates and the preset grid specification parameters, and the grid area parameters of each grid area are marked according to the diagonal coordinates.

[0042] As an example, Figure 3 A schematic diagram showing the gridding of parking spaces is shown in FIG. Figure 3 As shown in the figure, a spatial rectangular coordinate system is constructed, and the parking area is marked in the spatial rectangular coordinate system. The parking area rectangle ABCD is gridded according to the given grid size, and a set of diagonal coordinates Gs (x Gs ,y Gs )、Ge(x Ge ,y Ge ) to mark each grid area. Obtain the grid information of scattered scrap steel. Specifically, use the parameters Gs(x Gs ,y Gs )、Ge(x Ge ,y Ge ) Perform straight-through filtering on the scattered material point cloud in the parking area to extract the scattered material point cloud in each grid area, and traverse and calculate to obtain the maximum height value of the point cloud in each grid area, which is recorded as the height value h of the scattered material in the grid G , the center coordinates of each grid are The matching parameters of each grid area and the maximum height of scattered materials in the area (x G ,y G ,h G ) is the scattered material grid information of the parking area.

[0043] In one embodiment of the present application, calibrating the scanner calibration parameters includes: performing an initial scan of the target area to obtain three-dimensional point cloud data; placing a point cloud calibration marker in the target area, performing a secondary scan of the target area to obtain a scene point cloud containing the marker; and calculating the scanner calibration parameters through the three-dimensional point cloud data and the scene point cloud containing the marker.

[0044] Figure 4 FIG. 1 is a block diagram of a device for identifying scattered scrap steel according to an exemplary embodiment of the present application. Figure 4As shown, according to an embodiment of the present application, the scrap scattered material identification device 400 includes: an acquisition module 410, a training module 420, a data calibration module 430 and an identification module 440. The acquisition module is used to acquire scene point cloud samples, perform feature annotation on the scene point cloud samples, and obtain a training data set; the training module is used to perform training based on the selected neural network through the training data set to obtain a scrap scattered material point cloud segmentation model; the data calibration module is used to collect three-dimensional point cloud data of the target area, and calibrate the three-dimensional point cloud data to obtain three-dimensional point cloud data to be detected, the target area includes a designated parking area and a margin area around the parking area, and the margin area includes a space beyond a preset range of the boundary of the designated parking area; the identification module is used to input the three-dimensional point cloud data to be detected into the scrap scattered material point cloud segmentation model, obtain the scattered material point cloud of the target area, and calculate the scattered material grid information of the target area through the scattered material point cloud to complete the scrap scattered material identification. By acquiring scene point cloud samples and performing feature annotation, a training data set specifically for scattered scrap steel is constructed. The scattered scrap steel point cloud segmentation model trained based on this data set can more accurately identify scattered scrap steel in the target area, thereby improving the recognition accuracy and efficiency of scattered scrap steel. The entire recognition process is highly automated, from data collection, calibration to model recognition, which reduces manual intervention and improves work efficiency. In addition, the present application can not only accurately extract the location information of all scattered materials, including those that are smaller and difficult to separate from the ground, but also accurately measure their height information, ensuring the comprehensiveness and accuracy of the scattered material stacking operation, but also greatly improving the safety of the operation and avoiding potential risks caused by missing or misjudging scattered materials. In addition, the present application can also change the path planning and operation strategy that helps optimize unmanned vehicles, further reduce energy consumption and improve overall operation efficiency.

[0045] In one embodiment of the present application, the acquisition module is used to collect a scattered material point cloud sample set, which includes a variety of scattered material forms with different distributions; irregularly plot along the edge of each scattered material pile, assign a first label value to the scattered material, and assign a second label value to the non-scattered material point, to obtain a training data set

[0046] In one embodiment of the present application, a data calibration module is used to perform three-dimensional laser scanning on a target area to obtain an initial point cloud of the target area; the initial point cloud data is calibrated and calibrated by using scanner calibration parameters to obtain calibrated point cloud data; the scanner calibration parameters are used to correct scanner installation errors; the calibrated point cloud data is pass-filtered according to a parking area threshold to filter out point clouds in non-parking areas to obtain three-dimensional point cloud data to be detected.

[0047] In one embodiment of the present application, an identification module is used to use a scattered material segmentation model to deduce the three-dimensional point cloud data to be detected to obtain a deduction result, wherein the deduction result includes a label value for each point in the three-dimensional point cloud data to be detected in a specified parking area; traverse each point and its label value in the deduction result of the scattered material segmentation model, and synthesize the points with a label value into a scattered material point cloud.

[0048] In one embodiment of the present application, the scrap scattered material identification device also includes a grid information calculation module 450, which is used to grid the scattered material point cloud in the target area according to preset grid specification parameters to obtain multiple grid areas, and the grid area includes grid area parameters; the scattered material point cloud in the target area is directly filtered through the grid area parameters to obtain a grid scattered point cloud; the point cloud height values ​​of each grid scattered point cloud are traversed, and the maximum height value point in each grid scattered point cloud is recorded as the grid maximum height value point; the grid area parameters and the grid maximum height value point are determined as the scattered material grid information of the target area.

[0049] In one embodiment of the present application, the calculation module is used to establish a spatial rectangular coordinate system if the target area is a parking space area with a rectangular shape, and mark the vertex corners of the parking area to obtain the vertex coordinates of the parking area; grid the marked parking space area according to preset grid specification parameters to obtain multiple grid areas, calculate the diagonal coordinates of the grid area according to the vertex coordinates and the preset grid specification parameters, and mark the grid area parameters of each grid area according to the diagonal coordinates.

[0050] In one embodiment of the present application, a data calibration module is used to perform an initial scan of a target area to obtain three-dimensional point cloud data; a point cloud calibration marker is placed in the target area, and a secondary scan is performed on the target area to obtain a scene point cloud containing the marker; and the scanner calibration parameters are calculated using the three-dimensional point cloud data and the scene point cloud containing the marker.

[0051] It should be noted that the device provided in the above embodiment and the method provided in the above embodiment belong to the same concept, wherein the specific manner in which each module and unit performs the operation has been described in detail in the method embodiment, and will not be repeated here. In practical applications, the device provided in the above embodiment can distribute the above functions to different functional modules as needed, that is, divide the internal structure of the device into different functional modules to complete all or part of the functions described above, and this is not limited here.

[0052] See also Figure 5 , Figure 5 is a schematic diagram of a structure of an electronic device shown in an exemplary embodiment of the present application. It should be noted that: Figure 5The electronic device 500 shown is only an example and should not bring any limitation to the functions and usage area of ​​the embodiments of the present application.

[0053] like Figure 5 As shown, the electronic device 500 includes a processor 501, a memory 502 and a communication bus 503; the communication bus 503 is used to connect the processor 501 and the memory 502; the processor 501 is used to execute the computer program stored in the memory 502 to implement one or more methods in the above embodiments.

[0054] The electronic device provided in the present application includes a processor, a memory, a transceiver and a communication interface. The memory and the communication interface are connected to the processor and the transceiver and complete communication with each other. The memory is used to store computer programs, the communication interface is used to communicate, and the processor and the transceiver are used to run the computer program so that the electronic device executes each step of the above method.

[0055] In this embodiment, the memory may include a random access memory (RAM), and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0056] The above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0057] The above embodiments are merely illustrative of the principles and effects of the present application, and are not intended to limit the present application. Anyone familiar with the technology may modify or change the above embodiments without violating the spirit and scope of the present application. Therefore, all equivalent modifications or changes made by a person of ordinary skill in the art without departing from the spirit and technical ideas disclosed in the present application shall still be covered by the claims of the present application.

Claims

1. A method for identifying scattered scrap steel, characterized in that: The method comprises: Acquire scene point cloud samples, perform feature annotation on the scene point cloud samples, and obtain a training data set; Through the training data set, training is performed based on the selected neural network to obtain a point cloud segmentation model for scattered scrap steel; Collecting three-dimensional point cloud data of a target area, and calibrating the three-dimensional point cloud data to obtain three-dimensional point cloud data to be detected, wherein the target area includes a designated parking area and a margin area around the parking area, and the margin area includes a space within a preset range beyond the boundary of the designated parking area; The three-dimensional point cloud data to be detected is input into the scattered scrap material point cloud segmentation model to obtain the scattered material point cloud of the target area, and the scattered material grid information of the target area is calculated by the scattered material point cloud to complete the scattered scrap material identification.

2. The method for identifying scattered scrap steel according to claim 1, characterized in that: Obtain scene point cloud samples, perform feature annotation on the scene point cloud samples, and obtain a training data set, including: Collecting a scattered material point cloud sample set, wherein the scattered material point cloud sample set includes a plurality of scattered material forms with different distributions; Irregular plotting is performed along the edge of each scattered material pile, the scattered material is assigned a first label value, and the non-scattered material point is assigned a second label value to obtain a training data set.

3. The method for identifying scattered scrap steel according to claim 1, characterized in that: Collect the 3D point cloud data of the target area and calibrate the point cloud to obtain the 3D point cloud data to be detected, including: Performing three-dimensional laser scanning on the target area to obtain an initial point cloud of the target area; Calibrate and calibrate the initial point cloud data using scanner calibration parameters to obtain calibrated point cloud data; the scanner calibration parameters are used to correct scanner installation errors; The calibrated point cloud data is subjected to through-filtering according to the parking space area threshold, and the point cloud of the non-parking space area is filtered out to obtain the three-dimensional point cloud data to be detected.

4. The method for identifying scattered scrap steel according to claim 1, characterized in that: The three-dimensional point cloud data to be detected is input into the scrap scattered material point cloud segmentation model to obtain the scattered material point cloud of the target area, including: The three-dimensional point cloud data to be detected are deduced using the scattered material segmentation model to obtain a deduction result, wherein the deduction result includes a label value of each point in each three-dimensional point cloud data to be detected in the designated parking area; Traverse each point and its label value in the deduction result of the scattered material segmentation model, and synthesize the points with a label value into a scattered material point cloud.

5. The method for identifying scattered scrap steel according to claim 1, characterized in that: The scattered material grid information of the target area is calculated by computing the scattered material point cloud, and further includes: Gridding the scattered material point cloud in the target area according to preset grid specification parameters to obtain a plurality of grid areas, wherein the grid areas include grid area parameters; Performing through-filter processing on the scattered material point cloud in the target area by using the grid area parameters to obtain a grid scattered point cloud; Traverse the point cloud height values ​​of each grid scattered point cloud, and record the maximum height value point in each grid scattered point cloud as the grid maximum height value point; The grid area parameters and the grid maximum height value point are determined as the scattered material grid information of the target area.

6. The method for identifying scattered scrap steel according to claim 5, characterized in that: The scattered material point cloud in the target area is gridded according to preset grid specification parameters to obtain a plurality of grid areas, wherein the grid areas include grid area parameters, including: If the target area is a parking space area with a rectangular shape, a spatial rectangular coordinate system is established, and the vertex angles of the parking area are marked to obtain the vertex coordinates of the parking area; The marked parking space area is grid-divided according to preset grid specification parameters to obtain multiple grid areas, the diagonal coordinates of the grid areas are calculated according to the vertex coordinates and the preset grid specification parameters, and the grid area parameters of each grid area are marked according to the diagonal coordinates.

7. The method for identifying scattered scrap according to claim 3, characterized in that: Calibrate the scanner calibration parameters, including: Performing an initial scan on the target area to obtain three-dimensional point cloud data; Placing a point cloud calibration marker in a target area, performing a second scan on the target area, and obtaining a scene point cloud containing the marker; The scanner calibration parameters are calculated by using the three-dimensional point cloud data and the scene point cloud containing the marker.

8. A device for identifying scattered scrap steel, characterized in that: The device comprises: An acquisition module is used to acquire scene point cloud samples, perform feature annotation on the scene point cloud samples, and obtain a training data set; A training module, used for training based on the selected neural network through the training data set to obtain a point cloud segmentation model for scattered scrap steel; A data calibration module is used to collect three-dimensional point cloud data of a target area and calibrate the three-dimensional point cloud data to obtain three-dimensional point cloud data to be detected, wherein the target area includes a designated parking area and a margin area around the parking area, and the margin area includes a space within a preset range beyond the boundary of the designated parking area; The recognition module is used to input the three-dimensional point cloud data to be detected into the scrap steel scattered material point cloud segmentation model to obtain the scattered material point cloud of the target area, and calculate the scattered material grid information of the target area through the scattered material point cloud to complete the scrap steel scattered material recognition.

9. An electronic device, characterized in that: The electronic device comprises: one or more processors; A storage device for storing one or more programs, which, when executed by the one or more processors, enables the electronic device to implement the method for identifying scattered scrap as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and the computer program is used to enable a computer to execute the method for identifying scattered scrap steel as claimed in any one of claims 1 to 7.

Images