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 includes: collecting first point cloud data around the track; eliminating points in the first point cloud data that are far from the ground and far from the track to obtain second point cloud data containing track data; converting the second point cloud data to obtain a depth image; extracting lines from the depth image; filtering the lines based on a preset line matching restriction rule according to the positional relationship between two tracks to obtain the lines of the track. This method can improve the efficiency of track identification.

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, in the field of track detection, PCA (principal component analysis) can be used to extract the track direction and obtain the track characteristics. In this way, the point cloud data of the track collected at a close distance has no obvious abnormalities. However, in long-distance PCA, when the track has a small inclination angle, due to the reflection of the track surface, it may be impossible to collect point cloud data at a long distance. In this case, the extraction of track characteristics will fail. Summary of the invention

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

[0004] One or more embodiments of the present invention provide a track identification method, comprising: collecting first point cloud data around the track; eliminating points in the first point cloud data that are far from the ground and far from the track to obtain second point cloud data containing track data; converting the second point cloud data to obtain a depth image; extracting lines from the depth image; filtering the lines based on a preset line matching restriction rule according to the positional relationship between two tracks to obtain the lines of the track.

[0005] Optionally, the method further includes: after obtaining the line of the track, determining a track obstacle avoidance interval according to the line of the track.

[0006] Optionally, the method further includes: after obtaining a depth image based on the second point cloud data conversion, setting filtering parameters according to the characteristics of the track, wherein the filtering parameters include: a range of a filtering window is greater than a preset track width value; filtering the depth image according to the filtering parameters to obtain a filtered image; and subtracting the depth image from the filtered image to obtain a track image.

[0007] Optionally, the method further includes: binarizing the track image to obtain a binary image; performing edge extraction on the binary image to obtain an image after edge extraction; and extracting lines from the image after edge extraction by using a Hough transform algorithm.

[0008] Optionally, the preset line matching restriction rule includes:

[0009]

[0010] d0<Abs(Image2(xi1,yi1)-Image2(xi1,yi1))<d1;

[0011] d0<Abs(Image2(xi2,yi2)-Image2(xi2,yi2))<d1;

[0012] Among them, xi1 represents the horizontal coordinate of the starting point of the i-th track in the image, yi1 represents the vertical coordinate of the starting point of the i-th track in the image, xi2 represents the horizontal coordinate of the end point of the i-th line in the image, yi2 represents the vertical coordinate of the end point of the i-th line in the image, d0 represents the lower limit of the preset track width value, and d1 represents the upper limit of the preset track width value; according to the positional relationship between the two tracks, the lines are filtered based on the preset line matching restriction rules to obtain the lines of the tracks, including: the lines in which multiple points simultaneously satisfy the above relationship equations are the lines of the tracks.

[0013] Optionally, filtering the lines based on a preset line matching restriction rule according to the positional relationship between the two tracks to obtain the lines of the track includes: extracting the lines of the track from the lines satisfying any of the following relationship expressions;

[0014] Abs(y–y1(d0))≤m0;

[0015] Abs(y–y2(d0))≤m0;

[0016] Among them, y is the ordinate of a point on the track in the image, y1(d0) represents the ordinate of a point on a track in the image corresponding to when the track width value is d0, y2(d0) represents the ordinate of another point on the track in the image corresponding to when the track width value is d0, d0 represents the preset lower limit of the track width, and m0 represents the preset width value of the track line.

[0017] Optionally, determining the obstacle avoidance interval of the track according to the line of the track includes: determining the obstacle avoidance interval of the track from a position that satisfies the following two relations at the same time; y≥y1(d0)–m0; y≤y1(d0)+m0;

[0018] Among them, y is the vertical coordinate of the point on the track in the image, y1(d0) represents the vertical coordinate of a point on a track in the image corresponding to the track width value of d0, d0 represents the preset lower limit of the track width, and m0 represents the preset track line width value.

[0019] One or more embodiments of the present invention also provide a track identification device, comprising: an acquisition module, configured to acquire first point cloud data around the track; a removal module, configured to remove points in the first point cloud data that are far away from the ground and far away from the track, to obtain second point cloud data containing track data; a conversion module, configured to obtain a depth image based on the conversion of the second point cloud data; an extraction module, configured to extract lines from the depth image; and a filtering module, configured to filter the lines based on a preset line matching restriction rule according to the positional relationship between the two tracks, to obtain the lines of the track.

[0020] 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 the 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, and is used for any of the above-mentioned track identification methods.

[0021] 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.

[0022] The track identification method provided by one or more embodiments of the present invention collects first point cloud data around the track, eliminates points far away from the ground or the track, obtains second point cloud data, converts the second point cloud data to a two-dimensional projection plane, and obtains a depth image. Then, an image processing algorithm can be used to identify the lines of the track based on the obtained depth image, thereby improving the efficiency of track identification. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] 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.

[0024] Figure 1 is a flow chart of a track identification method according to one or more embodiments of the present invention;

[0025] FIG2( a) is a schematic diagram of an image before filtering according to one or more embodiments of the present invention;

[0026] FIG2( b ) is a schematic diagram of an image after filtering the image of FIG2( a ) according to one or more embodiments of the present invention;

[0027] FIG2(c) is a schematic diagram of an image obtained by subtracting FIG2(a) from FIG2(b);

[0028] FIG3( a) is a schematic diagram of an image before filtering according to one or more embodiments of the present invention;

[0029] FIG3( b ) is a schematic diagram of an image after filtering the image of FIG3( a ) according to one or more embodiments of the present invention;

[0030] FIG3(c) is a schematic diagram of an image obtained by subtracting FIG3(a) from FIG3(b) according to one or more embodiments of the present invention;

[0031] Figure 4 is a schematic diagram of a binary image obtained based on a depth map according to one or more embodiments of the present invention;

[0032] Figure 5 is a schematic diagram of an image after line extraction according to one or more embodiments of the present invention;

[0033] Figure 6 is a schematic diagram showing a trapezoidal feature extracted according to the characteristics of the track itself according to one or more embodiments of the present invention;

[0034] Figure 7 is a schematic diagram showing lines of extracted tracks according to one or more embodiments of the present invention;

[0035] Figure 8 is a schematic diagram showing an obstacle avoidance area of ​​an identified track according to one or more embodiments of the present invention;

[0036] Fig. 9 is a schematic structural diagram of a track identification device according to one or more embodiments of the present invention;

[0037] Fig.10 It is a schematic diagram showing the structure of an electronic device according to one or more embodiments of the present invention. DETAILED DESCRIPTION

[0038] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0039] 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.

[0040] 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:

[0041] Step 101: Collect first point cloud data around the track;

[0042] For example, an image acquisition device may be used to acquire first point cloud data, and the acquired first point cloud data may include the spatial position of each point in the space under the radar coordinate system. The laser radar has a longer ranging range and higher resolution in both horizontal and vertical directions, which can ensure that smaller targets can be detected.

[0043] Step 102: Eliminate points in the first point cloud data that are far from the ground and the track, to obtain second point cloud data containing track data;

[0044] For the first point cloud data collected in step 101, a ground extraction algorithm can be used to extract points with a normal vector of [0,0,1], and by setting a threshold, points far from the ground or the track [railway] are eliminated to obtain second point cloud data. The second point cloud data includes a point cloud of track data.

[0045] Step 103: Obtain a depth image according to the second point cloud data conversion;

[0046] The second point cloud data obtained in step 102 is converted into a depth image in the following manner:

[0047] Among them, the second point cloud data includes the three-dimensional coordinates (x, y, z) of the detected object, that is, each position around the track. According to the camera's field of view Fov and the radar's horizontal and vertical resolutions (u, v), it can be calculated:

[0048]

[0049] (u,v)=(arctan(y / x) / u', arcsin(z / d) / v');

[0050] Image1(u,v)=z;

[0051] Image2(u,v)=y;

[0052] Image3(u,v)=x;

[0053] In the above formula, D represents the distance value in the depth image, and x, y, and z represent the three-dimensional coordinates of a point.

[0054] Therefore, the distribution of the detected object in the image plane is continuous, and image-like information is obtained. The depth of the image is represented by the z value of the ground information.

[0055] Step 104: extracting lines from the depth image;

[0056] After obtaining the depth image, the depth image may be optimized. For example, the depth image may be filtered, opened, closed, and edge extracted to expand the point cloud information as much as possible to prevent the point cloud from being too little information and unable to extract lines. After optimizing the depth image, the lines in the image are extracted.

[0057] Step 105: filtering the lines based on the positional relationship between the two tracks and the preset line matching restriction rule to obtain the lines of the track.

[0058] Since there is a certain positional relationship between the two tracks to be detected, such as the width between the two tracks (hereinafter referred to as the track width), a line matching restriction rule can be established in advance based on the positional relationship. Only when the line extracted in step 104 satisfies the line matching restriction rule, the line can be determined as the line of the track.

[0059] The track identification method provided by one or more embodiments of the present invention collects first point cloud data around the track, eliminates points far away from the ground or the track, obtains second point cloud data, converts the second point cloud data to a two-dimensional projection plane, and obtains a depth image. Then, an image processing algorithm can be used to identify the lines of the track based on the obtained depth image, thereby improving the efficiency of track identification.

[0060] In one or more embodiments of the present invention, the track identification method may further include: after obtaining the track lines, determining the track obstacle avoidance interval according to the track lines. The area between two tracks may be the obstacle avoidance interval. Therefore, after extracting the track lines, the track danger avoidance interval of the point cloud, i.e., the three-dimensional danger avoidance space, may be obtained by back-projecting the image back into the point cloud.

[0061] In one or more embodiments of the present invention, the track identification method may further include:

[0062] After the depth image is obtained according to the conversion of the second point cloud data, the filtering parameters are set according to the characteristics of the track, wherein the filtering parameters include: the range of the filtering window is greater than the preset track width value; the depth image is filtered according to the filtering parameters to obtain a filtered image; the depth image and the filtered image are subtracted to obtain a track image. For example, according to the characteristics of the track, the filtering window can be set to exceed the track width, and the minimum value of the filtering window is output. In this way, two frames of image data can be obtained, and the filtered track data characteristics can be obtained by subtracting the two images. As shown in Figure 2(a), when the collected point cloud data of the track is relatively complete, after filtering, the data shown in 2(b) is obtained, and then when the two frames of images are subtracted, the image shown in 2(c) is obtained. The remaining part shown in 2(c) is the track, and the area between the two tracks is the obstacle avoidance area. As shown in Figure 3(a), when the point cloud data of the collected track is incomplete, the original image 3(a) is filtered to obtain image 3(b). After subtracting image 3(a) from image 3(b), the resulting image is shown in Figure 3(c). At this time, the image obtained is relatively sparse.

[0063] By comparing FIG. 2(c) and FIG. 3(c), it can be seen that when the point cloud data of the collected track is not complete, the track image obtained after filtering has poor quality. Therefore, in one or more embodiments of the present invention, after obtaining the track image, the track image may be optimized. Based on this, the track identification method of one or more embodiments of the present invention may further include: binarizing the track image to obtain a binary image; for example, binarizing, opening, closing, etc. may be performed on the track image according to the quality of the image to expand the information of the point cloud. The obtained binary image is as follows: Figure 4 After obtaining the binary image, edge extraction is performed on the binary image. For example, the Canny (edge ​​detection) operator can be used to perform edge detection to obtain an image after edge extraction. The lines in the image after edge extraction are extracted by the Hough transform algorithm. For the curves in the track, the image can be divided into different parts, and the straight line features in the image can be extracted in segments. The image after line extraction can be as follows Figure 5 As shown, Figure 5 The relatively continuous lines in the image are the lines of the track extracted by Hough transform, and the scattered points are other valid points remaining after filtering the image as described above.

[0064] like Figure 5 As shown, Figure 5 There is still a lot of noise in the lines, so these lines need to be further filtered to finally get the lines of the track.

[0065] The preset line matching restriction rule may include:

[0066]

[0067] d0<Abs(Im age2(xi1,yi1)-Im age2(xi1,yi1))<d1;

[0068] d0<Abs(Im age2(xi2,yi2)-Im age2(xi2,yi2))<d1;

[0069] Wherein, xi1 represents the abscissa of the starting point of the ith track in the image, yi1 represents the ordinate of the starting point of the ith track in the image, xi2 represents the abscissa of the end point of the ith line in the image, yi2 represents the ordinate of the end point of the ith line in the image, d0 represents the lower limit of the preset track width value, and d1 represents the upper limit of the preset track width value;

[0070] According to the positional relationship between the two tracks, the lines are filtered based on a preset line matching restriction rule to obtain the lines of the track, including:

[0071] The line in which multiple points in the line simultaneously satisfy the above relationship expressions is the line of the track.

[0072] For example, Figure 5 In the case of multiple lines in the image, it is necessary to use the positional relationship between the two tracks to perform line matching and extract the lines according to the characteristics of the tracks themselves. Figure 6 Trapezoidal features shown.

[0073] According to the actual characteristics of the track, the track should meet the following conditions in the image:

[0074] refer to Figure 6 As shown, the pixel difference between two lines at a close distance on the image is greater than the pixel difference at a long distance, and the angle between two lines at a close distance is greater than the angle between two lines at a long distance;

[0075] The spacing between two lines at different distances in the image meets the threshold range of the track width d. According to the actual characteristics of the track, the threshold range of d can be limited to the range of 1.2m to 2m. This width can be found by the position of the line in the image;

[0076] The height difference between the two tracks should be below a certain value;

[0077] Since there may be abnormal points in the image during linear extension, causing the linear part to be on some abnormal points, it is necessary to add spacing rules at multiple distances to avoid false rejection.

[0078] Still Figure 5 As an example, the line matching restriction rule is used to Figure 5 After filtering the lines in , the valid line interval can be obtained as Figure 7 As shown, the lines of the track are identified.

[0079] In one or more embodiments of the present invention, filtering the lines based on a preset line matching restriction rule according to the positional relationship between the two tracks to obtain the lines of the track may include:

[0080] Extracting the line of the track from the lines satisfying any of the following relationships;

[0081] Abs(y–y1(d0))≤m0;

[0082] Abs(y–y2(d0))≤m0;

[0083] Wherein, y is the ordinate of a point on the track in the image, y1(d0) represents the ordinate of a point on a track in the image corresponding to the track width value d0, y2(d0) represents the ordinate of a point on another track in the image corresponding to the track width value d0, d0 represents the preset lower limit of the track width, and m0 represents the preset track line width value. For example, according to the relationship between the two tracks constrained in the above equations, combined with the relationship between the x and y values, the limits of the y values ​​at each distance can be obtained, which are y1(d0), y2(d0), and the track line width m0 (indicates the width value of the track line in the image, which is smaller). When the depth value is d=d0, if the limit Abs(y–y1(d0))≤m0 or Abs(y–y2(d0)≤m0 is met, it is considered that the track line can be extracted.

[0084] In one or more embodiments of the present invention, determining the obstacle avoidance interval of the track according to the line of the track may include:

[0085] Determine the obstacle avoidance interval of the track from the positions that simultaneously satisfy the following two relations;

[0086] y ≥ y1(d0)–m0;

[0087] y≤y1(d0)+m0;

[0088] Wherein, y is the ordinate of the points on the track in the image, y1(d0) represents the ordinate of the points on a track in the image corresponding to the track width value of d0, d0 represents the preset lower limit of the track width, and m0 represents the preset width value of the track line. For example, when the constraint conditions: y≥y1(d0)–m0 and y≤y1(d0)+m0 are adopted, the obstacle avoidance area of the entire track can be obtained. At the same time, when the height at each distance satisfies the condition z<z0, where z0 represents the preset height threshold of the information value, there are no obstacles in the track area. In addition, it is also possible to determine whether there are obstacles or conventional objects such as wires in the track area according to the image captured by the camera. The schematic diagram of the obstacle avoidance area of the track extracted can be as Fig. 9 shown, wherein the part marked by the mark L is the obstacle avoidance area of the entire track.

[0089] Fig. 9 is a schematic structural diagram of an orbital recognition device shown according to one or more embodiments of the present invention, as Fig. 9 shown, the device 90 includes:

[0090] An acquisition module 91, configured to acquire first point cloud data around the track;

[0091] A rejection module 92, configured to reject the points in the first point cloud data that are far from the ground and far from the track, and obtain second point cloud data containing track data;

[0092] A conversion module 93, configured to convert a depth image according to the second point cloud data;

[0093] An extraction module 94, configured to extract lines from the depth image;

[0094] A filtering module 95, configured to filter the lines according to the positional relationship between two tracks based on a preset line matching constraint rule to obtain the lines of the track.

[0095] In one or more embodiments of the present invention, the above-mentioned orbital recognition device may further include: a determination module, configured to determine an orbital obstacle avoidance interval according to the lines of the track after obtaining the lines of the track.

[0096] In one or more embodiments of the present invention, the above-mentioned orbital recognition device may further include: a setting module, configured to set filtering parameters according to the characteristics of the track after converting a depth image according to the second point cloud data, wherein the filtering parameters include: the range of the filtering window is greater than the preset width value of the track;

[0097] A filtering module, configured to filter the depth image according to the filtering parameters to obtain a filtered image;

[0098] The first processing module is configured to subtract the depth image from the filtered image to obtain a track image.

[0099] In one or more embodiments of the present invention, the track identification device may further include: a second processing module configured to perform binarization processing on the track image to obtain a binary image;

[0100] A third processing module is configured to perform edge extraction processing on the binary image to obtain an image after edge extraction;

[0101] The fourth processing module is configured to extract lines in the image after edge extraction by using a Hough transform algorithm.

[0102] In one or more embodiments of the present invention, the preset line matching restriction rule may include:

[0103]

[0104] d0<Abs(Im age2(xi1,yi1)-Im age2(xi1,yi1))<d1;

[0105] d0<Abs(Im age2(xi2,yi2)-Im age2(xi2,yi2))<d1;

[0106] Wherein, xi1 represents the abscissa of the starting point of the ith track in the image, yi1 represents the ordinate of the starting point of the ith track in the image, xi2 represents the abscissa of the end point of the ith line in the image, yi2 represents the ordinate of the end point of the ith line in the image, d0 represents the lower limit of the preset track width value, and d1 represents the upper limit of the preset track width value;

[0107] According to the positional relationship between the two tracks, the lines are filtered based on a preset line matching restriction rule to obtain the lines of the track, including:

[0108] The line in which multiple points in the line simultaneously satisfy the above relationship expressions is the line of the track.

[0109] The filtering module is specifically configured as follows:

[0110] Extracting the line of the track from the lines satisfying any of the following relationships;

[0111] Abs(y–y1(d0))≤m0;

[0112] Abs(y–y2(d0))≤m0;

[0113] Among them, y is the ordinate of a point on the track in the image, y1(d0) represents the ordinate of a point on a track in the image corresponding to when the track width value is d0, y2(d0) represents the ordinate of another point on the track in the image corresponding to when the track width value is d0, d0 represents the preset lower limit of the track width, and m0 represents the preset width value of the track line.

[0114] The determination module is specifically configured as follows:

[0115] Determine the obstacle avoidance interval of the track from the positions that simultaneously satisfy the following two relations;

[0116] y ≥ y1(d0)–m0;

[0117] y≤y1(d0)+m0;

[0118] Among them, y is the vertical coordinate of the point on the track in the image, y1(d0) represents the vertical coordinate of a point on a track in the image corresponding to the track width value of d0, d0 represents the preset lower limit of the track width, and m0 represents the preset track line width value.

[0119] 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.

[0120] 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.

[0121] Correspondingly, such as Fig.10 As shown, the electronic device provided by one or more embodiments of the present invention may include: a housing 11, a processor 12, a memory 13, a circuit board 14 and a power supply circuit 15, wherein the circuit board 14 is arranged inside the space enclosed by the housing 11, and the processor 12 and the memory 13 are arranged on the circuit board 14; the power supply circuit 15 is used to supply power to various circuits or devices of the electronic device; the memory 13 is used to store executable program codes; the processor 12 runs a program corresponding to the executable program code by reading the executable program code stored in the memory 13, so as to execute any one of the track identification methods provided by the aforementioned embodiments.

[0122] 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.

[0123] 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.

[0124] 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.

[0125] 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.

[0126] 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.

[0127] 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: Collecting first point cloud data around the track; removing points in the first point cloud data that are far from the ground and the track, to obtain second point cloud data containing track data; Obtain a depth image according to the second point cloud data conversion; extract lines from the depth image; filter the lines based on a preset line matching restriction rule according to the positional relationship between the two tracks to obtain the lines of the track; After obtaining the line of the track, determining the track obstacle avoidance interval according to the line of the track; After obtaining a depth image according to the second point cloud data conversion, filter parameters are set according to the characteristics of the track, wherein the filter parameters include: the range of the filter window is greater than the preset track width value; the depth image is filtered according to the filter parameters to obtain a filtered image; the depth image and the filtered image are subtracted to obtain a track image; the track image is binarized to obtain a binary image; the binary image is edge extracted to obtain an edge extracted image; and lines in the edge extracted image are extracted by a Hough transform algorithm.

2. The method according to claim 1, characterized in that The preset line matching restriction rules include: 0<(y22-y12) / (y21-y11)<1; d0<Abs(Image2(xi1,yi1)-Image2(xi1,yi1))<d1; d0<Abs(Image2(xi2,yi2)-Image2(xi2,yi2))<d1; Among them, xi1 represents the horizontal coordinate of the starting point of the i-th track in the image, yi1 represents the vertical coordinate of the starting point of the i-th track in the image, xi2 represents the horizontal coordinate of the end point of the i-th line in the image, yi2 represents the vertical coordinate of the end point of the i-th line in the image, d0 represents the lower limit of the preset track width value, and d1 represents the upper limit of the preset track width value; according to the positional relationship between the two tracks, the lines are filtered based on the preset line matching restriction rules to obtain the lines of the tracks, including: the lines in which multiple points simultaneously meet the preset line matching restriction rules are the lines of the tracks.

3. The method according to claim 1, characterized in that The lines are filtered based on a preset line matching restriction rule according to the positional relationship between the two tracks to obtain the lines of the track, including: extracting the lines of the track from the lines that satisfy any of the following relationship expressions; Abs(y–y1(d0))≤m0; Abs(y–y2(d0))≤m0; Among them, y is the ordinate of a point on the track in the image, y1(d0) represents the ordinate of a point on a track in the image corresponding to when the track width value is d0, y2(d0) represents the ordinate of another point on the track in the image corresponding to when the track width value is d0, d0 represents the preset lower limit of the track width, and m0 represents the preset width value of the track line.

4. The method according to claim 2, characterized in that: Determining an obstacle avoidance interval of the track according to the lines of the track includes: Determine the obstacle avoidance interval of the track from the positions that simultaneously satisfy the following two relations; y≥y1(d0)–m0; y≤y1(d0)+m0; Among them, y is the vertical coordinate of the point on the track in the image, y1(d0) represents the vertical coordinate of a point on a track in the image corresponding to the track width value of d0, d0 represents the preset lower limit of the track width, and m0 represents the preset track line width value.

5. A track identification device, characterized in that: include: A collection module is configured to collect first point cloud data around the track; A rejection module is configured to reject points far from the ground and the track in the first point cloud data to obtain second point cloud data containing track data; a conversion module is configured to obtain a depth image according to the conversion of the second point cloud data; An extraction module, configured to extract lines from the depth image; A filtering module, configured to filter the lines based on a preset line matching restriction rule according to the positional relationship between the two tracks to obtain the lines of the track; After obtaining the line of the track, determining the track obstacle avoidance interval according to the line of the track; After obtaining a depth image according to the second point cloud data conversion, filter parameters are set according to the characteristics of the track, wherein the filter parameters include: the range of the filter window is greater than the preset track width value; the depth image is filtered according to the filter parameters to obtain a filtered image; the depth image and the filtered image are subtracted to obtain a track image; the track image is binarized to obtain a binary image; the binary image is edge extracted to obtain an edge extracted image; and lines in the edge extracted image are extracted by a Hough transform algorithm.

6. 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 4.

7. 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 4.

Images