A Detection Method, System and Medium for the Distance from the Center Line of the Coupler to the Rail Surface

Abstract

The present invention discloses a detection method, system and medium for the distance from the coupler center line to the rail surface; it relates to the technical field of distance measurement; it is improved on the basis of traditional measurement technologies. The camera of the inspection robot is used to collect the coupler image and the rail image at the same point, and the coupler image and the rail image are respectively subjected to mask segmentation and projection to obtain the coupler center line point cloud and the rail surface point cloud; after the coupler center line point cloud and the rail surface point cloud are converted to the base coordinate system of the inspection robot based on the transformation matrix, the centroid point coordinates of the coupler center line point cloud and the rail plane equation are obtained; it realizes the rapid and accurate measurement of the distance from the coupler center line to the rail surface, saves a large amount of human resource investment, and can avoid measurement errors caused by different levels of manual business.

Description

Technical Field

[0001] The present invention relates to the technical field of distance measurement, and particularly to a detection method, system and medium for the distance from the coupler center line to the rail surface. Background Art

[0002] The height of the coupler center line from the rail surface is the vertical distance between the center line of the hook connecting the train carriages and the railway track. In railway transportation, the height of the coupler center line from the rail surface is a key parameter, which plays an important role in ensuring the stability and safety of railway trains. The adjustment of the height of the coupler center line from the rail surface directly affects the traction and braking force of the vehicle, and also affects the stability and safety of the vehicle. When the height of the coupler center line from the rail surface is too large or too small, it will cause lateral and longitudinal torsional forces when the train passes through curves or slopes, and even lead to derailment. Therefore, precise control of the height of the coupler center line from the rail surface is very important. The traditional measurement method is to place the measuring ruler base on both sides of the rail, make the cross beam span across the rail, lock and fix the main ruler after adjusting the position, and select the probe positioning method according to the coupler of the EMU. The value directly read from the reading device after the probe is aligned with the measurement position is the coupler center height at this time. The traditional measurement method has complex steps and consumes a large amount of manpower. Summary of the Invention

[0003] The technical problem to be solved by the present invention is that the traditional measurement method is to place the measuring ruler base on both sides of the rail, make the cross beam span across the rail, lock and fix the main ruler after adjusting the position, and select the probe positioning method according to the coupler of the EMU, with complex steps and consuming a large amount of manpower. The purpose of the present invention is to provide a detection method, system and medium for the distance from the coupler center line to the rail surface, which is improved on the basis of traditional measurement technology, uses the camera of the inspection robot to collect the coupler image and the rail image at the same point, and performs corresponding computer operation processing on the coupler image and the rail image to realize rapid and accurate measurement of the distance from the coupler center line to the rail surface, saving a large amount of human resource investment, and avoiding measurement errors caused by different levels of manual business.

[0004] The present invention is achieved by the following technical solutions:

[0005] This solution provides a detection method for the distance from the coupler center line to the rail surface, including:

[0006] Performing hand-eye calibration on the camera of the inspection robot to obtain the transformation matrix from the camera coordinate system to the end effector of the inspection robot's manipulator;

[0007] Collecting the coupler image and the rail image at the same point based on the camera of the inspection robot;

[0008] Performing mask segmentation and projection on the coupler image and the rail image respectively to obtain the coupler center line point cloud and the rail surface point cloud;

[0009] Convert the coupler center line point cloud and the track surface point cloud to the inspection robot base coordinate system based on the transformation matrix, extract the centroid point coordinates based on the coupler center line point cloud, and fit the track plane equation based on the track surface point cloud;

[0010] Calculate the distance from the centroid point coordinates to the track plane equation as the distance from the coupler center line to the rail surface.

[0011] The working principle of this solution: The traditional measurement method is to place the measuring scale base on both sides of the rail, make the crossbeam span across the rail, lock and fix it after adjusting the position of the main scale, and select the probe positioning method according to the coupler of the EMU. The steps are complex and consume a large amount of manpower; the purpose of the present invention is to provide a detection method, system and medium for the distance from the coupler center line to the rail surface, which is improved on the basis of the traditional measurement technology, uses the camera of the inspection robot to collect the coupler image and the track image at the same point, and performs corresponding computer operation processing on the coupler image and the track image to realize accurate measurement of the distance from the coupler center line to the rail surface quickly, saving a large amount of human resource investment, and can avoid measurement errors caused by different artificial business levels.

[0012] A further optimized solution is to perform hand-eye calibration on the camera of the inspection robot to obtain the transformation matrix from the camera coordinate system to the end effector of the inspection robot manipulator, including the method:

[0013] Install the camera on the end effector of the inspection robot manipulator to obtain the transformation matrix for converting the camera coordinate system to the end effector coordinate system of the manipulator

[0014] A further optimized solution is to perform mask segmentation and projection on the coupler image and the track image respectively to obtain the coupler center line point cloud and the track center line point cloud; including the method:

[0015] Extract the coupler center line mask image mask of the coupler image based on the PIDNET deep learning semantic segmentation network CG , project the coupler center line mask image mask CG onto the point cloud to obtain the coupler center line point cloud

[0016] Extract the track surface mask image mask of the track image based on the PIDNET deep learning semantic segmentation network gd ; project the track mask image mask gd onto the point cloud to obtain the track surface point cloud

[0017] A further optimized solution is that the PIDNET deep learning semantic segmentation network includes a proportional real-time semantic segmentation branch, an integral semantic segmentation branch, and a derivative semantic segmentation branch;

[0018] The proportional real-time semantic segmentation branch is used to parse and save the detailed information in the high-resolution feature map;

[0019] The integral real-time semantic segmentation branch is used to summarize local and global context information to parse long-range dependencies;

[0020] The derivative real-time semantic segmentation branch is used to extract high-frequency features to predict the boundary region;

[0021] The loss function of the PIDNET deep learning semantic segmentation network is:

[0022] Loss=λ0l0+λ1l1+λ2l2+λ3l3

[0023] Where λ0, λ1, λ2, and λ3 are training hyperparameters, which are set to 0.4, 20, 1, and 1 respectively; l0 represents the semantic loss, l1 represents the weighted binary cross-entropy loss, and l2 and l3 represent the cross-entropy loss respectively.

[0024] A further optimized solution is that the method for converting the coupler centerline point cloud and the track surface point cloud to the inspection robot base coordinate system based on the transformation matrix includes:

[0025] Converting the coupler centerline point cloud to the inspection robot base coordinate system based on the transformation matrix has:

[0026]

[0027] Where, represents the coupler centerline point cloud in the inspection robot base coordinate system; is the inherent transformation matrix for converting the coordinates of the end effector of the inspection robot manipulator to the base coordinate system;

[0028] Converting the track surface point cloud to the base coordinate system based on the transformation matrix has:

[0029]

[0030] Where, represents the track surface point cloud in the inspection robot base coordinate system.

[0031] A further optimized solution is that the method for extracting the centroid point coordinates based on the coupler centerline point cloud includes:

[0032] Input the coupler centerline points into the following formula to extract the centroid point coordinates:

[0033]

[0034] where r i =(x i , y i , z i ), i = 1, 2, 3,..., n represents the point cloud on the coupler center line, and n represents the total number of point clouds; m i is the calculation weight of point cloud i; M is the total number of point clouds on the coupler center line; Pc(x c , y c , z c ) is the calculated centroid point coordinate.

[0035] A further optimization solution is that the track plane equation is fitted based on the track plane point cloud; the method includes:

[0036] Obtain the track plane point cloud and construct the track plane equation model as:

[0037] Ax + By + Cz + D = 0

[0038] where A, B, and C respectively represent the components of the plane normal vector, and D represents the offset from the origin;

[0039] Convert the track plane equation model to:

[0040] z = Ax + By + C

[0041] Construct the following error function s for fitting. When the error reaches the minimum, the track plane equation is obtained; the error function s is expressed as:

[0042]

[0043] where: (x i , y i , z i ), i = 0, 1..n are the track plane point cloud data.

[0044] A further optimization solution is that the distance from the calculated centroid point coordinate to the track plane equation is used as the distance from the coupler center line to the rail surface, and the method includes:

[0045] Calculate the distance d from the coupler center line to the rail surface according to the following formula:

[0046]

[0047] where x c , y c , z c represents the centroid coordinate of the coupler center line.

[0048] The present solution also provides a detection system for the distance from the coupler center line to the rail surface, which is used to implement the detection method for the distance from the coupler center line to the rail surface as described above. The system includes:

[0049] A calibration module, which is used to perform hand-eye calibration on the camera of the inspection robot to obtain the transformation matrix from the camera coordinate system to the end effector of the robotic arm of the inspection robot;

[0050] An acquisition module, which is used to acquire the coupler image and the track image at the same point based on the camera of the inspection robot;

[0051] A segmentation and projection module, which is used to perform mask segmentation and projection on the coupler image and the track image respectively to obtain the coupler center line point cloud and the track plane point cloud;

[0052] A fitting module, which is used to transform the coupler center line point cloud and the track plane point cloud to the base coordinate system of the inspection robot based on the transformation matrix, extract the centroid point coordinates based on the coupler center line point cloud, and fit the track plane equation based on the track plane point cloud;

[0053] A calculation module, which is used to calculate the distance from the centroid point coordinates to the track plane equation as the distance from the coupler center line to the rail surface.

[0054] The present solution also provides a computer-readable medium, on which a computer program is stored. The computer program, when executed by a processor, can implement the detection method for the distance from the coupler center line to the rail surface as described above.

[0055] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0056] The present invention provides a detection method, system and medium for the distance from the coupler center line to the rail surface. It is improved on the basis of traditional measurement techniques. The camera of the inspection robot is used to acquire the coupler image and the track image at the same point, and corresponding computer arithmetic processing is performed on the coupler image and the track image, so as to quickly and accurately measure the distance from the coupler center line to the rail surface, save a large amount of human resource investment, and avoid measurement errors caused by different levels of manual business. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the technical solutions of the exemplary embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention, and therefore should not be regarded as limiting the scope. For those of ordinary skill in the art, other related drawings can be obtained based on these drawings without creative efforts. In the drawings:

[0058] Figure 1 It is a schematic flowchart of the detection method for the distance from the coupler center line to the rail surface;

[0059] Figure 2 Schematic diagram of the detection principle for the distance from the coupler center line to the rail surface;

[0060] Figure 3 Mask image of the coupler center line;

[0061] Figure 4 Schematic diagram of the coupler center line point cloud;

[0062] Figure 5 Mask image of the rail plane;

[0063] Figure 6 Schematic diagram of the rail plane point cloud;

[0064] Figure 7 Schematic diagram of converting the coupler center line point cloud and the rail plane point cloud to the robot base coordinate system. Detailed implementation manners

[0065] To make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below in conjunction with embodiments and drawings. The illustrative embodiments and descriptions of the present invention are only used to explain the present invention and are not intended to limit the present invention.

[0066] The traditional measurement method is to place the measuring scale seat on both sides of the rail, make the cross beam span across the rail, lock and fix it after adjusting the position of the main scale, and select the probe positioning method according to the coupler of the EMU. The steps are complex and consume a large amount of manpower. In view of this, the following embodiments are provided in this solution to solve the above technical problems.

[0067] Embodiment 1

[0068] This embodiment provides a method for detecting the distance from the coupler center line to the rail surface, as shown in Figure 1 and Figure 2 , including:

[0069] Step 1: Perform hand-eye calibration on the inspection robot camera to obtain the transformation matrix from the camera coordinate system to the end effector of the inspection robot manipulator; this step specifically includes the method:

[0070] Install the camera on the end effector of the inspection robot manipulator to obtain the transformation matrix for converting the camera coordinate system to the end effector coordinate system of the manipulator

[0071] Step 2: Collect the coupler image and the rail image based on the inspection robot camera at the same point;

[0072] Step 3: Perform mask segmentation and projection on the coupler image and the rail image respectively to obtain the coupler center line point cloud and the rail surface point cloud; this step specifically includes the method:

[0073] Based on the PIDNET deep learning semantic segmentation network, the coupler center line mask image mask of the coupler image is extracted. CG , as Figure 3 shown, Figure 3 The left side is the 2D image of the coupler, and the right side is the mask image of the coupler center line; project the coupler center line mask image mask CG onto the point cloud to obtain the coupler center line point cloud as Figure 4 shown, Figure 4 The black part in the middle on the left side is the point cloud of the coupler center line, and the right side is the enlarged view of the extracted point cloud.

[0074] as Figure 5 shown, based on the PIDNET deep learning semantic segmentation network, the track plane mask image mask of the track image is extracted gd ; Figure 5 The left side is the captured track image, and the right side is the extracted track plane mask; as Figure 6 shown, project the track mask image mask gd onto the point cloud to obtain the track point cloud

[0075] The PIDNET deep learning semantic segmentation network includes a proportional real-time semantic segmentation branch, an integral semantic segmentation branch, and a derivative semantic segmentation branch;

[0076] The proportional real-time semantic segmentation branch is used to parse and save the detailed information in the high-resolution feature map;

[0077] The integral real-time semantic segmentation branch is used to summarize local and global context information to parse long-range dependencies;

[0078] The derivative real-time semantic segmentation branch is used to extract high-frequency features to predict the boundary region;

[0079] The loss function of the PIDNET deep learning semantic segmentation network is:

[0080] Loss = λ0l0 + λ1l1 + λ2l2 + λ3l3

[0081] where λ0, λ1, λ2, λ3 are training hyperparameters, which are set to 0.4, 20, 1, 1 respectively; l0 represents the semantic loss, l1 represents the weighted binary cross-entropy loss, and l2 and l3 represent the cross-entropy loss respectively.

[0082] Step four, as Figure 7 shown, based on the transformation matrix, the coupler center line point cloud and the track point cloud are transformed into the base coordinate system of the inspection robot. Based on the coupler center line point cloud, the centroid point coordinates are extracted, and based on the track point cloud, the track plane equation is fitted;Figure 7 The upper figure shows the point cloud of the coupler center line, and the lower figure shows the point cloud of the track plane;

[0083] The conversion of the coupler center line point cloud and the track plane point cloud to the base coordinate system of the inspection robot based on the transformation matrix includes the following method:

[0084] The conversion of the coupler center line point cloud to the base coordinate system of the inspection robot based on the transformation matrix is as follows:

[0085]

[0086] Among them, represents the coupler center line point cloud in the base coordinate system of the inspection robot; is the inherent transformation matrix for converting the coordinates of the end effector of the inspection robot's robotic arm to the base coordinate system;

[0087] The conversion of the track plane point cloud to the base coordinate system based on the transformation matrix is as follows:

[0088]

[0089] Among them, represents the track plane point cloud in the base coordinate system of the inspection robot.

[0090] The extraction of the centroid point coordinates based on the coupler center line point cloud includes the following method:

[0091] Input the coupler center line points into the following formula to extract the centroid point coordinates:

[0092]

[0093] where r i =(x i , y i , z i ) i = 1, 2, 3,..., n represents the points on the coupler center line, and n represents the total number of points; m i is the calculation weight of point cloud i; M is the total number of points in the coupler center line point cloud; Pc(x c , y c , z c ) is the calculated centroid point coordinates.

[0094] The fitting of the track plane equation based on the track plane point cloud; includes the following method:

[0095] Obtain the track plane point cloud and construct the track plane equation model as:

[0096] Ax + By + Cz + D = 0

[0097] Among them, A, B, and C respectively represent the components of the plane normal vector, and D represents the offset from the origin;

[0098] Convert the orbital plane equation model to:

[0099] z = Ax + By + C

[0100] Construct the following error function s for fitting. When the error reaches the minimum, the orbital plane equation is obtained. The error function s is expressed as:

[0101]

[0102] where: (x i , y i , z i )i = 0, 1..n are the orbital plane point cloud data.

[0103] Step 5, calculate the distance from the centroid point coordinates to the orbital plane equation as the distance from the coupler center line to the rail surface.

[0104] This step specifically includes the method:

[0105] Calculate the distance d from the coupler center line to the rail surface according to the following formula:

[0106]

[0107] where, x c , y c , z c represent the centroid coordinates of the coupler center line.

[0108] Since the coupler and the orbital plane have a large field of view, the current 3D camera cannot capture the coupler and the orbital plane in one image at the same time. This solution innovatively proposes to separately capture the coupler and the rail surface at one point, and use hand-eye calibration to convert them into the robot base coordinate system for accurate measurement from the coupler to the rail surface in the same coordinate system. The specific description is as follows: A color structured light 3D camera is built at the end of the robotic arm of the robot. The camera collects RGB-D images, and the RGB-D images include 3D point clouds and 2D images. After the robot reaches the shooting point, use the robotic arm to teach the pose, and horizontally shoot the coupler and vertically shoot the upper surface of the track respectively to obtain the RGB-D images of the coupler and the track. Perform deep learning semantic segmentation on the 2D image of the coupler, segment the coupler center line, then map it to the 3D point cloud, and extract the center line point cloud. Perform deep learning semantic segmentation on the 2D color image of the track, segment the 2D image of the track, then map it to the 3D point cloud, and extract the 3D point cloud of the track. Finally, convert the extracted 3D point cloud of the coupler center line and the extracted 3D point cloud of the track into the robot reference coordinate system, so that the coupler center line point cloud and the track point cloud are in the same coordinate system. After converting to the same coordinate system, fit a plane to the track point cloud and calculate the distance from the coupler center line to the orbital plane, which is the height of the coupler center line from the rail surface.

[0109] The present invention utilizes the 3D camera of the inspection robot to quickly and accurately measure the coupler, saving a large amount of human resource investment and avoiding measurement errors caused by different levels of manual business skills.

[0110] Embodiment 2

[0111] This embodiment provides a detection system for the distance from the coupler center line to the rail surface, which is used to implement a detection method for the distance from the coupler center line to the rail surface as described in Embodiment 1. The system includes:

[0112] A calibration module for performing hand-eye calibration on the inspection robot camera to obtain the transformation matrix from the camera coordinate system to the end effector of the inspection robot manipulator;

[0113] An acquisition module for acquiring coupler images and rail images at the same point based on the inspection robot camera;

[0114] A segmentation and projection module for performing mask segmentation and projection on the coupler image and the rail image respectively to obtain the coupler center line point cloud and the rail surface point cloud;

[0115] A fitting module for converting the coupler center line point cloud and the rail surface point cloud to the base coordinate system of the inspection robot based on the transformation matrix, extracting the centroid point coordinates based on the coupler center line point cloud, and fitting the rail plane equation based on the rail surface point cloud;

[0116] A calculation module for calculating the distance from the centroid point coordinates to the rail plane equation as the distance from the coupler center line to the rail surface.

[0117] Embodiment 3

[0118] This embodiment provides a computer-readable medium, on which a computer program is stored. The computer program, when executed by a processor, can implement a detection method for the distance from the coupler center line to the rail surface as described in Embodiment 1; specifically, the following steps are executed:

[0119] Step 1: Perform hand-eye calibration on the inspection robot camera to obtain the transformation matrix from the camera coordinate system to the end effector of the inspection robot manipulator;

[0120] Step 2: Acquire coupler images and rail images at the same point based on the inspection robot camera;

[0121] Step 3: Perform mask segmentation and projection on the coupler image and the rail image respectively to obtain the coupler center line point cloud and the rail surface point cloud;

[0122] Step 4: Transform the coupler centerline point cloud and the track surface point cloud to the base coordinate system of the inspection robot based on the transformation matrix. Extract the centroid point coordinates based on the coupler centerline point cloud, and fit the track plane equation based on the track surface point cloud;

[0123] Step 5: Calculate the distance from the centroid point coordinates to the track plane equation as the distance from the coupler centerline to the rail surface.

[0124] The specific embodiments described above further elaborate on the purpose, technical solutions, and beneficial effects of the present invention. It should be understood that the above are only specific embodiments of the present invention and are not used to limit the protection scope of the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A detection method for the distance from the center line of the coupler to the rail surface, characterized in that Including: Perform hand-eye calibration on the inspection robot camera to obtain the transformation matrix from the camera coordinate system to the end effector of the inspection robot's manipulator; Collect coupler images and track images based on the inspection robot camera at the same position; Mask segmentation and projection are performed on the coupler image and the track image respectively to obtain the coupler centerline point cloud and the track surface point cloud; specifically, the method includes: extracting the coupler centerline mask image of the coupler image based on the PIDNET deep learning semantic segmentation network , and projecting the coupler centerline mask image onto the point cloud to obtain the coupler centerline point cloud ; extracting the track surface mask image of the track image based on the PIDNET deep learning semantic segmentation network ; projecting the track mask image onto the point cloud to obtain the track surface point cloud ; The PIDNET deep learning semantic segmentation network includes a proportional real-time semantic segmentation branch, an integral real-time semantic segmentation branch, and a derivative real-time semantic segmentation branch; the proportional real-time semantic segmentation branch is used to parse and save the detailed information in the high-resolution feature map; the integral real-time semantic segmentation branch is used to summarize local and global context information to parse long-range dependencies; the derivative real-time semantic segmentation branch is used to extract high-frequency features to predict the boundary region; the loss function of the PIDNET deep learning semantic segmentation network is: ; where , , , are training hyperparameters, which are set to 0.4, 20, 1, 1 respectively; represents the semantic loss, represents the weighted binary cross-entropy loss, and respectively represent the cross-entropy loss; Transform the coupler centerline point cloud and the track point cloud to the base coordinate system of the inspection robot based on the transformation matrix, extract the centroid point coordinates based on the coupler centerline point cloud, and fit the track plane equation based on the track point cloud; Calculate the distance from the centroid point coordinates to the track plane equation as the distance from the coupler centerline to the rail surface.

2. The detection method of the distance from the center line of the coupler to the rail surface according to claim 1, characterized in that, The method for performing hand-eye calibration on the inspection robot camera to obtain the transformation matrix from the camera coordinate system to the end effector of the inspection robot's manipulator includes: Install the camera on the end effector of the inspection robot manipulator, and obtain the transformation matrix for converting the camera coordinate system to the end effector coordinate system of the manipulator .

3. The detection method of the distance from the center line of the coupler to the rail surface according to claim 1, wherein The method for transforming the coupler centerline point cloud and the track point cloud to the base coordinate system of the inspection robot based on the transformation matrix includes: The methods for transforming the coupler centerline point cloud to the base coordinate system of the inspection robot based on the transformation matrix are: ； Among them, represents the point cloud of the coupler center line in the base coordinate system of the inspection robot; is the inherent transformation matrix for converting the coordinates of the end effector of the inspection robot's manipulator to the base coordinate system; The methods for transforming the track point cloud to the base coordinate system are: ； Among them, represents the point cloud of the track surface in the base coordinate system of the inspection robot.

4. The detection method of the distance from the coupler center line to the rail surface according to claim 3, characterized in that, The method for extracting the centroid point coordinates based on the coupler centerline point cloud includes: Input the coupler centerline points into the following formula to extract the centroid point coordinates: ； Among them represents the point cloud on the center line of the coupler; i = 1, 2, 3,..., n; n represents the total number of point clouds; is the calculation weight of point cloud i; M is the total number of point clouds on the center line of the coupler.

5. The detection method of the distance from the center line of the coupler to the rail surface according to claim 3, characterized in that, Fitting the track plane equation based on the track point cloud; Including methods: Obtain the track point cloud and construct the track plane equation model as: Where A, B, and C respectively represent the components of the plane normal vector, and D represents the offset from the origin; Convert the track plane equation model to: z = Ax + By + C Construct the following error function s for fitting, and when the error reaches the minimum, obtain the track plane equation; the error function s is expressed as: Wherein: is the orbital plane point cloud data, and i = 0, 1..n.

6. The detection method for the distance from the coupler center line to the rail surface according to claim 5, wherein The method for calculating the distance from the centroid point coordinates to the track plane equation as the distance from the coupler centerline to the rail surface includes: Calculate the distance d from the coupler centerline to the rail surface according to the following formula: Among them, x c , y c , z c represent the centroid coordinates of the coupler center line.

7. A detection system for the distance from the center line of a coupler to the rail surface, characterized in that, For implementing a method for detecting the distance from the coupler centerline to the rail surface according to any one of claims 1-6, the system includes: A calibration module for performing hand-eye calibration on the inspection robot camera to obtain the transformation matrix from the camera coordinate system to the end effector of the inspection robot's manipulator; An acquisition module for collecting coupler images and track images based on the inspection robot camera at the same position; A segmentation and projection module for performing mask segmentation and projection on the coupler image and the track image respectively to obtain the coupler centerline point cloud and the track point cloud; A fitting module for transforming the coupler centerline point cloud and the track point cloud to the base coordinate system of the inspection robot based on the transformation matrix, extracting the centroid point coordinates based on the coupler centerline point cloud, and fitting the track plane equation based on the track point cloud; A calculation module is used to calculate the distance from the centroid point coordinates to the orbital plane equation as the distance from the coupler center line to the rail surface.

8. A computer-readable medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it can implement a method for detecting the distance from the coupler center line to the rail surface according to any one of claims 1-6.

Images