Track identification method, device, storage medium and equipment

Abstract

One or more embodiments of the present invention provide a track identification method, apparatus, storage medium and device, wherein the track identification method may include: acquiring first point cloud data around the track; converting the first point cloud data to an X-Y plane; establishing a first model based on the geometric features of the converted first point cloud data; comparing the first model with a preset track model to determine a target point cloud in the first point cloud data that matches the preset track model, wherein the preset track model is established based on the geometric features of the point cloud of the track; identifying the track based on the target point cloud can improve the track identification accuracy.

Description

Technical Field

[0001] The present invention relates to the field of image processing technology, and in particular to a track recognition method, device, storage medium and equipment. Background Art

[0002] At present, the track extraction based on vision cannot reflect the three-dimensional track information, and it is difficult to identify the extracted track in weak light conditions. The track extraction method based on three-dimensional point cloud is usually based on edge features. It will first calculate the normal vector and extract the edge based on the normal vector. In this way, more edge line features will be extracted. For tracks at a long distance, the track extraction error is usually large due to a change in the slope of a point. Therefore, a track recognition method needs to be proposed. Summary of the invention

[0003] In view of this, one or more embodiments of the present invention provide a track identification method, apparatus, storage medium and device, which can effectively improve the track identification accuracy.

[0004] One or more embodiments of the present invention provide a track identification method, comprising: acquiring first point cloud data around a track; converting the first point cloud data to an XY plane; establishing a first model based on the geometric features of the converted first point cloud data; comparing the first model with a preset track model to determine a target point cloud in the first point cloud data that matches the preset track model, wherein the preset track model is established based on the geometric features of the point cloud of the track; and identifying the track based on the target point cloud.

[0005] Optionally, establishing a first model according to geometric features of the converted first point cloud data includes: dividing the converted first point cloud data into multiple grids; and establishing the first model corresponding to each grid in the multiple grids according to geometric features of each grid.

[0006] Optionally, the first model corresponding to each grid among the multiple grids is established according to the geometric features of each grid, including: calculating the maximum value of the point cloud in each grid on the Z axis; normalizing the maximum value of the point cloud in each grid on the Z axis to obtain the first model corresponding to the point cloud of each grid.

[0007] Optionally, the method also includes: before obtaining the first point cloud data around the track, obtaining second point cloud data, wherein each point cloud in the second point cloud data is a point cloud of the track; calculating the maximum value and minimum value of the second point cloud data on the Y axis, and the maximum value of the second point cloud data on the Z axis; dividing the second point cloud data in the Y axis direction into multiple intervals with a preset unit length according to the maximum value and minimum value of the second point cloud data on the Y axis, and normalizing each interval by taking the maximum value of the point cloud in the interval on the Z axis to obtain the preset track model corresponding to the interval.

[0008] Optionally, identifying the track according to the target point cloud includes: determining a center point of the target point cloud; filtering the center point to obtain a filtered center point; and determining an expression for a center line of the track according to the filtered center point.

[0009] Optionally, the method further includes: after obtaining an expression for the center line of the track, determining an obstacle avoidance area of ​​the track according to the expression.

[0010] One or more embodiments of the present invention also provide a track identification device, comprising: a first acquisition module, configured to acquire first point cloud data around the track; a conversion module, configured to convert the first point cloud data to an XY plane; an establishment module, configured to establish a first model based on the geometric features of the converted first point cloud data; a first determination module, configured to compare the first model with a preset track model, and determine a target point cloud in the first point cloud data that matches the preset track model, wherein the preset track model is established based on the geometric features of the point cloud of the track; and an identification module, configured to identify the track based on the target point cloud.

[0011] Optionally, the establishing module is specifically configured to: divide the converted first point cloud data into a plurality of grids; and establish the first model corresponding to each grid according to geometric features of each grid in the plurality of grids.

[0012] One or more embodiments of the present invention further provide an electronic device, comprising: a housing, a processor, a memory, a circuit board and a power supply circuit, wherein the circuit board is placed inside a space enclosed by the housing, and the processor and the memory are arranged on the circuit board; the power supply circuit is used to supply power to various circuits or devices of the electronic device; the memory is used to store executable program code; the processor runs a program corresponding to the executable program code by reading the executable program code stored in the memory, so as to execute any one of the above-mentioned track identification methods.

[0013] One or more embodiments of the present invention further provide a non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium stores computer instructions, wherein the computer instructions are used to enable the computer to execute any one of the above-mentioned track identification methods.

[0014] The track identification method of one or more embodiments of the present invention first obtains first point cloud data collected around the track, converts the point cloud data to an XY plane, establishes a first model based on the geometric features of the converted first point cloud data, and compares the first model with a preset track model pre-established based on the geometric features of the point cloud of the track to determine a target point cloud that matches the preset track model. Since the target point cloud matches the preset track model, the track is identified based on the target point cloud, which can improve the track identification accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0016] Figure 1 The figure is a flow chart showing a track identification method according to one or more embodiments of the present invention.

[0017] Figure 2 is a schematic diagram of a preset track model according to one or more embodiments of the present invention.

[0018] Figure 3 is a schematic diagram of a track area obtained through prediction according to one or more embodiments of the present invention.

[0019] Figure 4 According to one or more embodiments of the present invention, Figure 3 A schematic diagram of the track area obtained after radius filtering of the point cloud in the track area is shown.

[0020] Figure 5 FIG. 4 is a schematic diagram showing a center line of a track obtained by linear fitting according to one or more embodiments of the present invention.

[0021] Figure 6 It is a schematic diagram of the structure of a track identification device according to one or more embodiments of the present invention.

[0022] Figure 7 It is a schematic diagram showing the structure of an electronic device according to one or more embodiments of the present invention. DETAILED DESCRIPTION

[0023] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0024] It should be clear that the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0025] Figure 1 is a flow chart of a track identification method according to one or more embodiments of the present invention. Figure 1 As shown, the method includes:

[0026] Step 101: Acquire first point cloud data around the track;

[0027] For example, point cloud data may be extracted block by block when the laser radar detects the track as the first point cloud data mentioned above, or point cloud data of the entire track may be obtained after the laser radar completes the detection of the track as the first point cloud data mentioned above.

[0028] Step 102: converting the first point cloud data to an XY plane;

[0029] For example, a plane fitting may be performed based on the first point cloud data to obtain a normal vector of the plane, and the normal vector is rotated to a (0, 0, 1) vector, thereby converting the first point cloud data to an XY plane.

[0030] Step 103: establishing a first model according to the geometric features of the converted first point cloud data;

[0031] For example, if in the above step 101, all point cloud data around the entire track are obtained as the first point cloud data after the laser radar completes the detection of the track, then in the above step 103, the converted point cloud data can be divided into multiple parts, and then the above first model is established based on the geometric features of the point cloud of each part in turn to obtain the first model corresponding to the point cloud of each part.

[0032] If in the above step 101, point clouds are extracted block by block when the laser radar collects point cloud data around the track, the point cloud data extracted each time can be used as the above first point cloud data, then in the above step 103, the above first model can be established based on the geometric features of the converted point cloud data.

[0033] Step 104: Compare the first model with a preset track model to determine a target point cloud in the first point cloud data that matches the preset track model;

[0034] When the first point cloud data is extracted block by block in the above step 101, the first model established based on the geometric features of the extracted first point cloud data can be compared with the pre-set preset track model each time to determine whether the currently collected first point cloud data belongs to the track area.

[0035] When the first point cloud data in the above step 101 is the point cloud data around the entire track obtained after the complete detection of the track, the first point cloud data can be divided into blocks to obtain multiple point clouds, and the above first model is established based on the geometric features of each point cloud. The first model corresponding to each point cloud can be compared with the preset track model to determine whether each point cloud belongs to the track area based on the similarity of the geometric features between the two.

[0036] After executing the above step 104, in order to improve the accuracy of track recognition, the point cloud determined not to belong to the track area can be eliminated, and the point cloud preliminarily determined to belong to the track area can be used as the target point cloud for subsequent track recognition processing.

[0037] Step 105: Identify the track according to the target point cloud.

[0038] The track identification method of one or more embodiments of the present invention first obtains first point cloud data collected around the track, converts the point cloud data to an XY plane, establishes a first model based on the geometric features of the converted first point cloud data, and compares the first model with a preset track model pre-established based on the geometric features of the point cloud of the track to determine a target point cloud that matches the preset track model. Since the target point cloud matches the preset track model, the track is identified based on the target point cloud, which can improve the track identification accuracy.

[0039] In one or more embodiments of the present invention, when the acquired point cloud data around the entire track is used as the above-mentioned first point cloud data, after converting the first point cloud data to the XY plane, the converted first point cloud data can be divided into blocks, and then the geometric features of the point clouds in each block are extracted, and a geometric feature model is established based on the geometric features of the point clouds in each block, that is, the above-mentioned first model can be obtained, so that the track data can be identified in smaller units. Based on this, the first model is established according to the geometric features of the converted first point cloud data, which may include: dividing the converted first point cloud data into multiple grids (an example of the above-mentioned multiple point clouds); establishing the first model corresponding to each grid according to the geometric features of each grid in the multiple grids. Based on this, the comparison of the first model with the preset track model in the above-mentioned step 104 can be calculated based on the following formula:

[0040]

[0041] In the above formula, shape_scence i Represents the first model established based on the geometric features of the i-th grid in the first point cloud data, shape_model i It represents the preset track model established based on the geometric features of the i-th area in the above-mentioned second point cloud data. Match_value is used to characterize the similarity between the first model of the i-th grid in the converted first point cloud data and the preset track model of the i-th area in the second point cloud data. When the calculated match_value is less than a threshold, it is preliminarily determined that the point cloud in the i-th grid in the first point cloud data belongs to the point cloud in the track area. Otherwise, it is determined that the i-th grid in the first point cloud data does not belong to the point cloud in the track area.

[0042] In one or more embodiments of the present invention, the track identification method may further include: obtaining second point cloud data before obtaining the first point cloud data around the track, wherein each point cloud in the second point cloud data is a point cloud of the track; calculating the maximum value and minimum value of the second point cloud data on the Y axis, and the maximum value of the second point cloud data on the Z axis; after calculating the minimum value and minimum value of the second point cloud data on the Y axis, the entire interval range of the second point cloud data on the Y axis can be determined, and the entire interval range of the second point cloud data on the Y axis can be divided into the above-mentioned multiple intervals by a preset unit length. According to the maximum value and minimum value of the second point cloud data on the Y axis, the second point cloud data is divided into multiple intervals in the Y-axis direction with a preset unit length, and each of the intervals takes the maximum value of the point cloud in the interval on the Z axis for normalization to obtain the preset track model corresponding to the interval. For example, it can be divided into 6 intervals in the Y-axis direction (as an example of the above-mentioned multiple intervals), and in each interval, the maximum value of the point cloud in the Z axis in the interval is taken and normalized. The preset track model obtained can be, for example, as follows Figure 2 As shown in the figure, BIN1-BIN6 represent 6 different intervals, BIN1 and BIN6 represent two track edges, which are somewhat raised relative to the ground, and BIN2-BIN5 represent the ground at the center of the track, which is lower than the track edge.

[0043] In one or more embodiments of the present invention, establishing the above-mentioned first model based on the first point cloud data is consistent with the manner of establishing the preset track model based on the second point cloud data described above. Based on this, establishing the first model corresponding to each grid in the multiple grids according to the geometric features of each grid may include: calculating the maximum value of the point cloud in each grid on the Z axis; normalizing the maximum value of the point cloud in each grid on the Z axis to obtain the first model corresponding to the point cloud of each grid.

[0044] In one or more embodiments of the present invention, identifying the track according to the target point cloud may include: calculating the center point of the target point cloud; for example, calculating the center point of the target point cloud predicted to belong to the track area, and determining the center point of the target point cloud predicted to belong to the track area after this step. Figure 3 As shown, Figure 3 The point cloud in the S area is the point cloud predicted to be the track, which may contain some point clouds that are mistakenly judged to belong to the track area. The center point is filtered to obtain the filtered center point; the expression of the center line of the track is determined according to the filtered center point. For example, the center point of the target point cloud can be radius filtered to eliminate the influence of the point cloud that is mistakenly judged to belong to the track area on the subsequent point cloud data fitting. Figure 3 Based on the area of ​​the track shown, the distribution of the center point after radius filtering can be shown as Figure 4 As shown in the figure, the point marked as G is the center point of the track area. The discrete point cloud predicted to be the track has been filtered out in the radius filtering process. Therefore, Figure 4 The center point of the track area marked in does not include these point clouds. Based on the filtered center point, the expression of the center line of the track can be obtained by the linear fitting formula, such as Figure 5 As shown, the line marked as L is the line in the image that satisfies the center line formula of the orbit obtained by fitting the filtered center point. Figure 5 It can be seen that the center line of the track obtained by linear fitting is consistent with the actual track, and there is no large error even at a long distance.

[0045] In one or more embodiments of the present invention, the track extraction method may further include: after obtaining the expression of the center line of the track, determining the obstacle avoidance area of ​​the track according to the expression. For example, judging whether the area to be detected is within the obstacle avoidance area of ​​the track according to the expression of the center line of the track, thereby realizing the detection of the obstacle avoidance object.

[0046] Figure 6 is a schematic diagram of a track identification device according to one or more embodiments of the present invention. Figure 6 As shown, the device 60 includes:

[0047] A first acquisition module 61 is configured to acquire first point cloud data around the track;

[0048] A conversion module 62, configured to convert the first point cloud data into an XY plane;

[0049] A building module 63 is configured to build a first model according to the geometric features of the converted first point cloud data;

[0050] A first determination module 64 is configured to compare the first model with a preset track model to determine a target point cloud in the first point cloud data that matches the preset track model, wherein the preset track model is established based on geometric features of the point cloud of the track;

[0051] The identification module 65 is configured to identify the track according to the target point cloud.

[0052] In one or more embodiments of the present invention, the establishment module may be specifically configured to: divide the converted first point cloud data into a plurality of grids; and establish the first model corresponding to each grid according to geometric features of each grid in the plurality of grids.

[0053] In one or more embodiments of the present invention, the establishment module is specifically configured to: normalize the maximum value of the point cloud in each grid on the Z axis to obtain the first model corresponding to the point cloud of each grid.

[0054] In one or more embodiments of the present invention, the track identification device may further include:

[0055] A second acquisition module is configured to acquire second point cloud data before acquiring first point cloud data around the track, wherein each point cloud in the second point cloud data is a point cloud of the track;

[0056] a calculation module, configured to calculate a maximum value and a minimum value of the second point cloud data on the Y axis, and a maximum value of the second point cloud data on the Z axis;

[0057] The division module is configured to divide the second point cloud data in the Y-axis direction into multiple intervals with a preset unit length according to the maximum value and the minimum value of the second point cloud data in the Y-axis, and normalize each interval by taking the maximum value of the point cloud in the Z-axis within the interval to obtain the preset track model corresponding to the interval.

[0058] In one or more embodiments of the present invention, the above-mentioned recognition module is specifically configured to: determine the center point of the target point cloud; filter the center point to obtain a filtered center point; and determine an expression of the center line of the track based on the filtered center point.

[0059] In one or more embodiments of the present invention, the track identification device may further include: a second determination module configured to determine the obstacle avoidance area of ​​the track according to the expression after obtaining the expression of the center line of the track.

[0060] One or more embodiments of the present invention further provide an electronic device, comprising: a housing, a processor, a memory, a circuit board and a power supply circuit, wherein the circuit board is placed inside a space enclosed by the housing, and the processor and the memory are arranged on the circuit board; the power supply circuit is used to supply power to various circuits or devices of the electronic device; the memory is used to store executable program code; the processor runs a program corresponding to the executable program code by reading the executable program code stored in the memory, so as to execute any one of the above-mentioned track identification methods.

[0061] One or more embodiments of the present invention further provide a non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium stores computer instructions, wherein the computer instructions are used to enable the computer to execute any one of the above-mentioned track identification methods.

[0062] Correspondingly, such as Figure 7 As shown, the server provided by the embodiment of the present invention may include: a shell 71, a processor 72, a memory 73, a circuit board 74 and a power supply circuit 75, wherein the circuit board 74 is arranged inside the space enclosed by the shell 71, and the processor 72 and the memory 73 are arranged on the circuit board 74; the power supply circuit 75 is used to supply power to various circuits or devices of the server; the memory 73 is used to store executable program codes; the processor 72 runs a program corresponding to the executable program code by reading the executable program code stored in the memory 73, so as to execute any one of the track identification methods provided by the aforementioned embodiments.

[0063] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0064] Each embodiment in this specification is described in a related manner, and the same or similar parts between the embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

[0065] In particular, for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0066] For the convenience of description, the above device is described by dividing the functions into various units / modules. Of course, when implementing the present invention, the functions of each unit / module can be implemented in the same or multiple software and / or hardware.

[0067] A person skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium, and when the program is executed, it can include the processes of the embodiments of the above-mentioned methods. The storage medium can be a disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc.

[0068] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. A track identification method, characterized in that: include: Acquire the first point cloud data around the track; Converting the first point cloud data to an XY plane; Establishing a first model according to the geometric features of the converted first point cloud data; The first model is compared with a preset track model to determine a target point cloud in the first point cloud data that matches the preset track model, wherein the preset track model is established based on the geometric features of the point cloud of the track; the track is identified according to the target point cloud; before obtaining the first point cloud data around the track, second point cloud data is obtained, wherein each point cloud in the second point cloud data is a point cloud of the track; the maximum and minimum values ​​of the second point cloud data on the Y axis, as well as the maximum value of the second point cloud data on the Z axis, are calculated; according to the maximum and minimum values ​​of the second point cloud data on the Y axis, the second point cloud data is divided into a plurality of intervals in the Y axis direction with a preset unit length, and each interval is normalized by taking the maximum value of the point cloud in the interval on the Z axis to obtain the preset track model corresponding to the interval.

2. The method according to claim 1, characterized in that Establishing a first model according to the geometric features of the converted first point cloud data includes: dividing the converted first point cloud data into a plurality of grids; and establishing the first model corresponding to each grid according to the geometric features of each grid in the plurality of grids.

3. The method according to claim 2, characterized in that The first model corresponding to each grid among the multiple grids is established according to the geometric features of each grid, including: calculating the maximum value of the point cloud in each grid on the Z axis; normalizing the maximum value of the point cloud in each grid on the Z axis to obtain the first model corresponding to the point cloud of each grid.

4. The method according to claim 1, characterized in that: Identifying the track according to the target point cloud includes: determining the center point of the target point cloud; filtering the center point to obtain a filtered center point; and determining an expression for a center line of the track according to the filtered center point.

5. The method according to claim 4, characterized in that The method further includes: after obtaining an expression for the center line of the track, determining an obstacle avoidance area of ​​the track according to the expression.

6. A track identification device, characterized in that: include: A first acquisition module is configured to acquire first point cloud data around the track; A conversion module, configured to convert the first point cloud data into an XY plane; A building module configured to build a first model according to the geometric features of the converted first point cloud data; A first determination module is configured to compare the first model with a preset track model to determine a target point cloud in the first point cloud data that matches the preset track model, wherein the preset track model is established based on the geometric features of the point cloud of the track; an identification module is configured to identify the track based on the target point cloud; a second acquisition module is configured to acquire second point cloud data before acquiring the first point cloud data around the track, wherein each point cloud in the second point cloud data is a point cloud of the track; a calculation module is configured to calculate the maximum and minimum values ​​of the second point cloud data on the Y axis, and the maximum value of the second point cloud data on the Z axis; the second point cloud data is divided into a plurality of intervals in the Y axis direction according to the maximum and minimum values ​​of the second point cloud data on the Y axis with a preset unit length, and each of the intervals takes the maximum value of the point cloud in the interval on the Z axis for normalization to obtain the preset track model corresponding to the interval.

7. The device according to claim 6, characterized in that The establishment module is specifically configured to: divide the converted first point cloud data into a plurality of grids; and establish the first model corresponding to each grid according to the geometric features of each grid in the plurality of grids.

8. An electronic device, characterized in that: The electronic device comprises: a shell, a processor, a memory, a circuit board and a power supply circuit, wherein the circuit board is placed inside the space enclosed by the shell, and the processor and the memory are arranged on the circuit board; the power supply circuit is used to supply power to various circuits or devices of the electronic device; the memory is used to store executable program code; the processor runs a program corresponding to the executable program code by reading the executable program code stored in the memory, so as to execute the track identification method described in any one of claims 1 to 5.

9. A non-transitory computer-readable storage medium, characterized in that: The non-transitory computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute the track identification method according to any one of claims 1 to 5.

Images