Non-contact wheel-rail dynamic posture parameter online visual measurement method and system

Abstract

This invention provides a non-contact online visual measurement method and system for wheel-rail dynamic attitude parameters, belonging to the field of 3D point cloud visual measurement technology. Based on the principle of stereo vision measurement, it reconstructs a wheel-rail cross-sectional profile point cloud sequence from the acquired image sequence of the structured light-illuminated cross-sectional area on the outer side of the wheel and rail, according to calibration parameters. Based on the reconstructed wheel-rail cross-sectional profile point cloud sequence, it partitions and fits planes, calculates and decomposes the angles of the normal vectors of specific planes using multi-frame information, and obtains angle parameters. Based on the reconstructed wheel-rail cross-sectional profile point cloud sequence, it fits the wheel circle using inter-frame and geometric constraints to determine key points in the wheel-rail region and the rail reference plane, and calculates wheel-rail displacement parameters. This invention enables simple, accurate, and safe non-contact measurement of lateral displacement, heave, yaw angle, and roll angle of the wheel during train operation; it provides rapid, accurate, and reliable theoretical and technical support for non-contact measurement of wheel-rail motion attitude.

Description

Technical Field

[0001] This invention relates to the field of 3D point cloud visual measurement technology, specifically to a non-contact online visual measurement method and system for wheel-rail dynamic attitude parameters. Background Technology

[0002] Online visual measurement of non-contact wheel-rail dynamic attitude parameters is the core technology for wheel-rail attitude parameter detection, which is of great significance for monitoring the real-time operation of wheel-rail and ensuring the safety of rail transit.

[0003] This method projects structured light onto the surface of a target, captures images of the projected area using a camera, and then reconstructs the target in three dimensions using stereo vision measurement principles to obtain the target's phase profile point cloud. This allows for the calculation of relevant physical parameters. It is a non-contact measurement method widely used in 3D imaging, industrial product quality inspection, and rail flaw detection. Current technologies for capturing wheel-rail motion during train operation largely rely on sensors and manual judgment, which poses certain safety risks during the measurement process. Summary of the Invention

[0004] The purpose of this invention is to provide a non-contact online visual measurement method and system for wheel-rail dynamic attitude parameters, so as to solve at least one of the technical problems existing in the background art.

[0005] To achieve the above objectives, the present invention adopts the following technical solution:

[0006] On one hand, the present invention provides a non-contact online visual measurement method for wheel-rail dynamic attitude parameters, comprising:

[0007] Acquire image sequences of the cross-sectional region illuminated by structured light on the outer side of the wheel and rail;

[0008] For the acquired image sequence of the structured light illumination section area on the outer side of the wheel and rail, the wheel and rail section profile point cloud sequence is reconstructed based on the stereo vision measurement principle and calibration parameters.

[0009] Based on the reconstructed wheel-rail cross-section profile point cloud sequence, the plane is fitted in sections, and the angle of the normal vector of a specific plane is calculated and decomposed by combining multi-frame information to obtain the angle parameters.

[0010] Based on the reconstructed wheel-rail cross-sectional profile point cloud sequence, the wheel circle is fitted by combining inter-frame and geometric constraints to determine the key points in the wheel-rail region and the rail reference plane, and the wheel-rail displacement parameters are calculated.

[0011] Preferably, acquiring the image sequence of the structured light irradiated section area outside the wheel and rail includes: calibrating the camera and structured light sensor using a standard checkerboard calibration plate before acquiring the wheel and rail structured light images; triggering acquisition when the train passes to acquire the image sequence of the structured light irradiated section area outside the wheel and rail within the camera's field of view.

[0012] Preferably, the wheel-rail cross-sectional profile point cloud sequence is reconstructed based on the parameters obtained by calibrating the camera and structured light sensor using a standard checkerboard calibration plate.

[0013] Preferably, the angle parameters to be calculated include:

[0014] Define a coordinate system where the x-axis is perpendicular to the direction of the rail extension and points outward, the y-axis is parallel to the direction of the rail extension, and the z-axis is perpendicular to the ground and points upward.

[0015] Based on the measured point cloud data S at time t t The three planes along the z-axis from top to bottom are fitted in partitions, and the angle between the normal vectors of two adjacent planes is calculated. Combined with the information from the previous frame, the corrected angle is calculated.

[0016] The rotation angle around the z-axis is the head-up angle; the rotation angle around the y-axis is the roll angle; where the angle between the two corrected normal vectors is decomposed to obtain their respective head-up angle and roll angle.

[0017] Calculate the head-shake angle and roll angle at time t based on the decomposition results.

[0018] Preferably, the calculation of wheel-rail displacement parameters includes:

[0019] Define a coordinate system where the x-axis is perpendicular to the direction of the rail extension and points outward, the y-axis is parallel to the direction of the rail extension, and the z-axis is perpendicular to the ground and points upward.

[0020] Based on the measured point cloud data S at time t t Partition fitting of three planes from top to bottom along the z-axis, and calculation of the equation of plane P based on the normal vectors of the three planes;

[0021] Using a corner detection algorithm, feature points in the wheel region are found. A spatial circle containing these feature points is then fitted by combining multi-frame motion and geometric constraints. The constraints are as follows: 1) The wheel center has a finite range of motion along the z-axis; 2) The wheel radius has prior geometric information. A fitting circle is defined, and the point on the fitting circle closest to plane P is considered a key point in the wheel region.

[0022] The displacement along the z-axis is the buoyancy; the displacement along the x-axis is the lateral displacement. The buoyancy is obtained by calculating the Euclidean distance between the key points in the wheel area along the z-axis and the plane P, and the lateral displacement is obtained by calculating the Euclidean distance between the key points in the wheel area along the x-axis and the plane P.

[0023] Preferably, the angle α between the two normal vectors is corrected. t and β t The equation is:

[0024] in and The following are the fitting errors for the three planes, in order. This indicates the confidence level of the included angle in the current frame. This represents the mean confidence level of the first t-1 frames. This represents the average angle of the first t-1 frames;

[0025] Calculate the head-shaking angle λ at time t. t and roll angle μ t for:

[0026] in and All are weighting coefficients, with a default value of 1.

[0027] The equation for plane P is:

[0028] in, These are the normal vectors of the three planes, (x) i ,y i ,z i ) represents the spatial coordinates of the i-th intersection point.

[0029] Secondly, the present invention provides a non-contact online visual measurement system for wheel-rail dynamic attitude parameters, comprising:

[0030] The acquisition module is used to acquire image sequences of the structured light illumination section area outside the wheel and rail.

[0031] The module is used to reconstruct the wheel-rail cross-sectional profile point cloud sequence based on the stereo vision measurement principle and calibration parameters for the acquired image sequence of the structured light illumination section area outside the wheel and rail.

[0032] The first calculation module is used to partition and fit planes based on the reconstructed wheel-rail profile point cloud sequence, and calculate and decompose the angle of the normal vector of a specific plane by combining multi-frame information to obtain the angle parameters.

[0033] The second calculation module is used to fit the wheel circle based on the reconstructed wheel-rail cross-sectional profile point cloud sequence, combined with inter-frame and geometric constraints, to determine the key points of the wheel-rail region and the rail reference plane, and to calculate the wheel-rail displacement parameters.

[0034] Thirdly, the present invention provides a non-transitory computer-readable storage medium for storing computer instructions, which, when executed by a processor, implement the non-contact online visual measurement method for wheel-rail dynamic attitude parameters as described above.

[0035] Fourthly, the present invention provides a computer program product, including a computer program that, when run on one or more processors, is used to implement the non-contact online visual measurement method for wheel-rail dynamic attitude parameters as described above.

[0036] Fifthly, the present invention provides an electronic device, comprising: a processor, a memory, and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory, causing the electronic device to execute instructions to implement the non-contact online visual measurement method for wheel-rail dynamic attitude parameters as described above.

[0037] The beneficial effects of this invention are: it enables simple, accurate, and safe measurement of the lateral displacement, heave, yaw angle, and roll angle of wheels during train operation in a non-contact manner; and it provides fast, accurate, and reliable theoretical and technical support for non-contact measurement of wheel-rail motion posture.

[0038] The advantages of additional aspects of the invention will be set forth more clearly in the following description or will be learned by practice of the invention. Attached Figure Description

[0039] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0040] Figure 1 This is a schematic diagram of the online visual measurement method for non-contact dynamic wheel-rail attitude parameters according to an embodiment of the present invention.

[0041] Figure 2 Figure (a) and Figure (b) are schematic diagrams of the equipment deployment scheme according to an embodiment of the present invention. Figure (a) and Figure (b) are the side view and top view of the deployment, respectively.

[0042] Figure 3 This is a flowchart of the angle parameter calculation method described in an embodiment of the present invention. Taking a point cloud reconstructed based on a line structured light sensor as an example, the planes L1, L2, and L3 to be fitted in Figures (a) and (b) correspond one-to-one. Figure (a) is a schematic diagram of the wheel-track region, Figure (b) is a visualization of the reconstructed point cloud from a certain viewpoint, and Figure (c) is a schematic diagram of angle calculation (λ). t μ t (These are the head-shake angle and roll angle at time t, respectively).

[0043] Figure 4Figures (a) and (b) are schematic diagrams of key points in the wheel region (the closest points on the wheel to the rail) at different times, Figure (c) is a schematic diagram of the rail reference plane, and Figure (d) is a schematic diagram of displacement parameter calculation (Z). t X t (These represent the buoyancy and lateral displacement at time t, respectively).

[0044] Figure 5 This is a schematic diagram of the measurement parameters described in an embodiment of the present invention (λ). t μ t Z t X t (These represent the head-shaking angle, roll angle, heave, and lateral displacement at time t). Detailed Implementation

[0045] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0046] It will be understood by those skilled in the art that, unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0047] It should also be understood that terms such as those defined in general dictionaries should be understood to have meanings consistent with their meanings in the context of the prior art, and should not be interpreted in an idealized or overly formal sense unless defined as here.

[0048] Those skilled in the art will understand that, unless specifically stated otherwise, the singular forms “a,” “an,” “the,” and “the” used herein may also include the plural forms. It should be further understood that the term “comprising” as used in this specification means the presence of the stated features, integers, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, and / or groups thereof.

[0049] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of those different embodiments or examples.

[0050] To facilitate understanding of the present invention, the present invention will be further explained and described below with reference to the accompanying drawings and specific embodiments. However, the specific embodiments do not constitute a limitation on the embodiments of the present invention.

[0051] Those skilled in the art should understand that the accompanying drawings are merely schematic diagrams of embodiments, and the components in the drawings are not necessarily essential for implementing the present invention.

[0052] Example 1

[0053] This embodiment 1 provides a non-contact online visual measurement system for wheel-rail dynamic attitude parameters, including:

[0054] The acquisition module is used to acquire image sequences of the structured light illumination section area outside the wheel and rail.

[0055] The module is used to reconstruct the wheel-rail cross-sectional profile point cloud sequence based on the stereo vision measurement principle and calibration parameters for the acquired image sequence of the structured light illumination section area outside the wheel and rail.

[0056] The first calculation module is used to partition and fit planes based on the reconstructed wheel-rail profile point cloud sequence, and calculate and decompose the angle of the normal vector of a specific plane by combining multi-frame information to obtain the angle parameters.

[0057] The second calculation module is used to fit the wheel circle based on the reconstructed wheel-rail cross-sectional profile point cloud sequence, combined with inter-frame and geometric constraints, to determine the key points of the wheel-rail region and the rail reference plane, and to calculate the wheel-rail displacement parameters.

[0058] In this embodiment 1, the above-described system is used to realize a non-contact online visual measurement method for wheel-rail dynamic attitude parameters, including:

[0059] The acquisition module was used to acquire image sequences of the structured light illumination section region outside the wheel and rail.

[0060] Using the construction module, based on the principle of stereo vision measurement, the wheel-rail cross-sectional profile point cloud sequence is reconstructed according to the calibration parameters for the acquired image sequence of the structured light illumination section area on the outer side of the wheel and rail.

[0061] Using the first calculation module based on the reconstructed wheel-rail cross-section profile point cloud sequence, the plane is fitted in sections, and the angle of the normal vector of a specific plane is calculated and decomposed by combining multi-frame information to obtain the angle parameters.

[0062] Using the second calculation module based on the reconstructed wheel-rail cross-sectional profile point cloud sequence, combined with inter-frame and geometric constraints to fit the wheel circle, key points in the wheel-rail region and the rail reference plane are determined, and wheel-rail displacement parameters are calculated.

[0063] Acquiring an image sequence of the structured light irradiated section area outside the wheel and rail includes: calibrating the camera and structured light sensor using a standard checkerboard calibration board before acquiring the structured light images of the wheel and rail; triggering acquisition when the train passes to acquire an image sequence of the structured light irradiated section area outside the wheel and rail within the camera's field of view.

[0064] The wheel-rail cross-sectional profile point cloud sequence was reconstructed based on the parameters calibrated using a standard checkerboard calibration plate for the camera and structured light sensor.

[0065] The calculation of angle parameters includes:

[0066] Define a coordinate system where the x-axis is perpendicular to the direction of the rail extension and points outward, the y-axis is parallel to the direction of the rail extension, and the z-axis is perpendicular to the ground and points upward.

[0067] Based on the measured point cloud data S at time t t The three planes along the z-axis from top to bottom are fitted in partitions, and the angle between the normal vectors of two adjacent planes is calculated. Combined with the information from the previous frame, the corrected angle is calculated.

[0068] The rotation angle around the z-axis is the head-up angle; the rotation angle around the y-axis is the roll angle; where the angle between the two corrected normal vectors is decomposed to obtain their respective head-up angle and roll angle.

[0069] Calculate the head-shake angle and roll angle at time t based on the decomposition results.

[0070] The calculation of wheel-rail displacement parameters includes:

[0071] Define a coordinate system where the x-axis is perpendicular to the direction of the rail extension and points outward, the y-axis is parallel to the direction of the rail extension, and the z-axis is perpendicular to the ground and points upward.

[0072] Based on the measured point cloud data S at time t t Partition fitting of three planes from top to bottom along the z-axis, and calculation of the equation of plane P based on the normal vectors of the three planes;

[0073] Using a corner detection algorithm, feature points in the wheel region are found. A spatial circle containing these feature points is then fitted by combining multi-frame motion and geometric constraints. The constraints are as follows: 1) The wheel center has a finite range of motion along the z-axis; 2) The wheel radius has prior geometric information. A fitting circle is defined, and the point on the fitting circle closest to plane P is considered a key point in the wheel region.

[0074] The displacement along the z-axis is the buoyancy; the displacement along the x-axis is the lateral displacement. The buoyancy is obtained by calculating the Euclidean distance between the key points in the wheel area along the z-axis and the plane P, and the lateral displacement is obtained by calculating the Euclidean distance between the key points in the wheel area along the x-axis and the plane P.

[0075] Correct the angle α between the two normal vectors t and β t The equation is:

[0076] in and The following are the fitting errors for the three planes, in order. This indicates the confidence level of the included angle in the current frame. This represents the mean confidence level of the first t-1 frames. This represents the average angle of the first t-1 frames;

[0077] Calculate the head-shaking angle λ at time t. t and roll angle μ t for:

[0078] in and All are weighting coefficients, with a default value of 1.

[0079] The equation for plane P is:

[0080] in, These are the normal vectors of the three planes, (x) i ,y i ,z i ) represents the spatial coordinates of the i-th intersection point.

[0081] In this embodiment 1, a structured light sensor and a high-resolution industrial camera are deployed and fixed beside the track. Continuous image acquisition is triggered when a train passes, obtaining a sequence of images of the structured light-illuminated area outside the wheel-rail within the camera's field of view. Specifically, the process includes the following steps:

[0082] Step S11: For the design and deployment of the front-end acquisition system, the installation distance and fixing method of the camera and structured light sensor must be determined based on the equipment parameters such as the power of the structured light sensor, field of view, camera resolution, frame rate, exposure time, field of view, and depth of field, combined with the on-site measurement environment, to ensure that the acquired images include the rail head, part of the rail web, and part of the wheels. See the appendix for equipment deployment details. Figure 2 Specifically: the vertical distance between the camera lens and the rail should be ≥100cm, the vertical distance between the structured light sensor and the rail should be ≥50cm, the camera and structured light sensor should be 2cm below the rail plane, and the angle between the structured light sensor and the camera should be between 20 and 45 degrees. The camera frame rate and exposure time depend on the train's speed to ensure clear and blur-free moving images are obtained.

[0083] Step S12: Adjust the illumination angle, brightness and other parameters of the structured light sensor so that the structured light covers the wheel and rail area to be measured clearly.

[0084] Step S13: Adjust the shooting angle so that the camera can capture clear structured light images in static conditions; adjust the camera exposure time and frame rate so that it can capture clear image sequences of high-speed moving objects.

[0085] In this embodiment 1, an online stereo vision measurement device is used to dynamically reconstruct the wheel-track profile point cloud sequence for each frame. The online stereo vision measurement device includes a front-end acquisition system and a back-end data processing system. The front-end acquisition system consists of a structured light sensor, a high-resolution industrial camera, an acquisition trigger, and fixtures. The back-end data processing system consists of data processing and analysis software running on a workstation. The structured light sensor projects a light plane onto the object being measured, forming light rays with specific structural features, thereby reconstructing the surface profile of the object and further performing parameter measurements. The industrial camera captures an image of the object with a specific structured light pattern on its surface, which is then used for 3D reconstruction and measurement. The acquisition trigger functions similarly to pressing a camera shutter to take a picture. In practice, the trigger is operated by clicking the "Shoot" button in the front-end acquisition system. The acquisition trigger is a well-known device in the prior art and belongs to the synchronization control module of the Thousand-Eyed Wolf high-speed image data storage system. The working logic of the data processing and analysis software, taking line structured light as an example, is as follows: 1) Image preprocessing: Use existing denoising methods to remove image noise, and crop the region of interest in the image to save subsequent computation; 2) Extract the center point of the light stripes on the calibration plate surface, and further use the cross-ratio invariant method to perform line structured light plane calibration; 3) Extract the center point of the light stripes on the surface of the object to be measured, and further use the triangulation method to reconstruct the object profile; 4) Define and measure parameters.

[0086] Example 2

[0087] like Figure 1As shown in this embodiment 2, a non-contact online visual measurement method for wheel-rail dynamic attitude parameters is provided, which can more simply, safely and accurately monitor the attitude parameters of wheels and rails during train operation.

[0088] Specifically, the method described in this embodiment 2 includes the following steps:

[0089] Step S1: Deploy and fix the structured light sensor and high-resolution industrial camera beside the track;

[0090] Step S2: When the train passes, continuous acquisition is triggered to obtain an image sequence of the structured light illumination area outside the wheel and rail within the camera's field of view;

[0091] Step S3: Dynamically reconstruct the wheel-track profile point cloud sequence for each frame using an online stereo vision measurement device;

[0092] Step S4: Based on the acquired point cloud, partition and fit the plane, combine multi-frame information to calculate and decompose the angle of the normal vector of a specific plane, and obtain the angle parameters.

[0093] Step S5: Based on the acquired point cloud, fit the wheel circle by combining inter-frame and geometric constraints, determine the key points of the wheel-rail region and the rail reference plane, and calculate the wheel-rail displacement parameters.

[0094] The online stereo vision measurement device includes a front-end acquisition system and a back-end data processing system. The front-end acquisition system consists of a structured light sensor, a high-resolution industrial camera, an acquisition trigger, and fixtures. The back-end data processing system consists of data processing and analysis software running on a workstation.

[0095] Step S1 includes:

[0096] Step S11: For the design and deployment of the front-end acquisition system, the installation distance and fixing method of the camera and structured light sensor must be determined based on the equipment parameters such as the power of the structured light sensor, field of view, camera resolution, frame rate, exposure time, field of view, and depth of field, combined with the on-site measurement environment, to ensure that the acquired images include the rail head, part of the rail web, and part of the wheels. See the appendix for equipment deployment details. Figure 2 Specifically: the vertical distance between the camera lens and the rail should be ≥100cm, the vertical distance between the structured light sensor and the rail should be ≥50cm, the camera and structured light sensor should be 2cm below the rail plane, and the angle between the structured light sensor and the camera should be between 20 and 45 degrees. The camera frame rate and exposure time depend on the train's speed to ensure clear and blur-free moving images are obtained.

[0097] Step S12: Adjust the illumination angle, brightness and other parameters of the structured light sensor so that the structured light covers the wheel and rail area to be measured clearly.

[0098] Step S13: Adjust the shooting angle so that the camera can capture clear structured light images in static conditions; adjust the camera exposure time and frame rate so that it can capture clear image sequences of high-speed moving objects.

[0099] Step S2 includes:

[0100] Step S21: Before data acquisition, perform high-precision calibration of the camera and structured light sensor using a standard checkerboard calibration board.

[0101] Step S22: Perform dynamic image acquisition when the train passes by, and obtain an image sequence of the structured light illumination area outside the wheel and rail within the camera's field of view.

[0102] Step S3 includes:

[0103] Step S31: Generate a wheel-rail profile point cloud sequence using the principle of stereo vision measurement.

[0104] Step S4 includes:

[0105] Step S41: Define a coordinate system with the x-axis perpendicular to the direction of the rail extension and pointing outwards, the y-axis parallel to the direction of the rail extension, and the z-axis perpendicular to the ground and pointing upwards.

[0106] Step S42: Based on the measured point cloud data S at time t t The partitioning fits three planes; see attached diagram. Figure 3 The three planes along the Z-axis from top to bottom are designated L1, L2, and L3 (L1 is the outer surface of the wheel, L2 is the outer surface of the wheel rim, and L3 is the middle section of the rail web). Figure 3 (Already labeled), let their fitting errors be respectively and Calculate the angle α between the normal vectors of planes L1 and L3 respectively. t The angle β between the normal vectors of L2 and L3 t Based on the information from the previous frame, the included angle is corrected using the following formula, where... This indicates the confidence level of the included angle in the current frame. This represents the mean confidence level of the first t-1 frames. This represents the average angle of the first t-1 frames.

[0107]

[0108] Step S43: Define angle parameters, such as Figure 4 As shown, specifically: 1) the rotation angle around the z-axis is the head-up angle; 2) the rotation angle around the y-axis is the side-roll angle. Decompose α separately. t and β t Get the head-shaking angle Roll angle The head-shaking angle λ at time t is calculated using the following formula.t and roll angle μ t :

[0109]

[0110] Step S5 includes:

[0111] Step S51: Define a coordinate system with the x-axis perpendicular to the direction of the rail extension and pointing outwards, the y-axis parallel to the direction of the rail extension, and the z-axis perpendicular to the ground and pointing upwards.

[0112] Step S52: Based on the measured point cloud data S at time t t Three planes are fitted to the partition, and these three planes are denoted as R1, R2, and R3 from top to bottom along the Z-axis (R1 is the side of the rail head, R2 is the plane area below the side of the rail head and above the lower jaw of the rail head, and R3 is the middle section of the rail web). The normal vectors are as follows: and Define plane P as perpendicular to R3 and passing through the intersection of the side edge and bottom edge of the rail head. Calculate the equation of plane P based on the following formula, where... For the plane normal vector, (x) i ,y i ,z i ) represents the spatial coordinates of the i-th intersection point.

[0113]

[0114] Step S53: Using a corner detection algorithm, find feature points in the wheel region, and fit a spatial circle containing the feature points by combining multi-frame motion and geometric constraints. The constraints are as follows: 1) The center of the wheel circle has a finite range of motion in the z-axis direction; 2) The radius of the wheel circle has geometric prior information. Define the fitted circle as R. Find the point on R that is closest to plane P, and consider it as the key point k in the wheel region.

[0115] Step S54: Define displacement parameters, such as Figure 5 As shown, specifically: 1) displacement along the z-axis—buoyancy; 2) displacement along the x-axis—lateral displacement. The buoyancy Z is obtained by calculating the Euclidean distance between k and plane P along the z-axis. t The transverse displacement X is obtained by calculating the Euclidean distance between k and plane P in the x-axis direction. t .

[0116] In summary, in this embodiment 2, based on the aforementioned measurement method, the lateral displacement, heave, yaw angle, and roll angle of the wheels during train operation can be measured simply, accurately, and safely in a non-contact manner. This technology provides rapid, accurate, and reliable theoretical and technical support for non-contact measurement of wheel-rail motion posture.

[0117] Example 3

[0118] Embodiment 3 of the present invention provides a non-transitory computer-readable storage medium for storing computer instructions. When the computer instructions are executed by a processor, a non-contact online visual measurement method for wheel-rail dynamic attitude parameters is implemented. The method includes:

[0119] Acquire image sequences of the cross-sectional region illuminated by structured light on the outer side of the wheel and rail;

[0120] For the acquired image sequence of the structured light illumination section area on the outer side of the wheel and rail, the wheel and rail section profile point cloud sequence is reconstructed based on the stereo vision measurement principle and calibration parameters.

[0121] Based on the reconstructed wheel-rail cross-section profile point cloud sequence, the plane is fitted in sections, and the angle of the normal vector of a specific plane is calculated and decomposed by combining multi-frame information to obtain the angle parameters.

[0122] Based on the reconstructed wheel-rail cross-sectional profile point cloud sequence, the wheel circle is fitted by combining inter-frame and geometric constraints to determine the key points in the wheel-rail region and the rail reference plane, and the wheel-rail displacement parameters are calculated.

[0123] Example 4

[0124] Embodiment 4 of the present invention provides a computer program (product), including a computer program that, when run on one or more processors, is used to implement a non-contact online visual measurement method for wheel-rail dynamic attitude parameters. The method includes:

[0125] Acquire image sequences of the cross-sectional region illuminated by structured light on the outer side of the wheel and rail;

[0126] For the acquired image sequence of the structured light illumination section area on the outer side of the wheel and rail, the wheel and rail section profile point cloud sequence is reconstructed based on the stereo vision measurement principle and calibration parameters.

[0127] Based on the reconstructed wheel-rail cross-section profile point cloud sequence, the plane is fitted in sections, and the angle of the normal vector of a specific plane is calculated and decomposed by combining multi-frame information to obtain the angle parameters.

[0128] Based on the reconstructed wheel-rail cross-sectional profile point cloud sequence, the wheel circle is fitted by combining inter-frame and geometric constraints to determine the key points in the wheel-rail region and the rail reference plane, and the wheel-rail displacement parameters are calculated.

[0129] Example 5

[0130] Embodiment 5 of the present invention provides an electronic device, including: a processor, a memory, and a computer program; wherein, the processor is connected to the memory, and the computer program is stored in the memory. When the electronic device is running, the processor executes the computer program stored in the memory to cause the electronic device to execute instructions for implementing a non-contact online visual measurement method for wheel-rail dynamic attitude parameters, the method including:

[0131] Acquire image sequences of the cross-sectional region illuminated by structured light on the outer side of the wheel and rail;

[0132] For the acquired image sequence of the structured light illumination section area on the outer side of the wheel and rail, the wheel and rail section profile point cloud sequence is reconstructed based on the stereo vision measurement principle and calibration parameters.

[0133] Based on the reconstructed wheel-rail cross-section profile point cloud sequence, the plane is fitted in sections, and the angle of the normal vector of a specific plane is calculated and decomposed by combining multi-frame information to obtain the angle parameters.

[0134] Based on the reconstructed wheel-rail cross-sectional profile point cloud sequence, the wheel circle is fitted by combining inter-frame and geometric constraints to determine the key points in the wheel-rail region and the rail reference plane, and the wheel-rail displacement parameters are calculated.

[0135] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0136] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0137] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0138] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment, whereby a series of operational steps are performed to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0139] While the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that, based on the technical solutions disclosed in the present invention, various modifications or variations that can be made by those skilled in the art without creative effort should be included within the scope of protection of the present invention.

Claims

1. A non-contact online visual measurement method for wheel-rail dynamic attitude parameters, characterized in that, include: Acquire image sequences of the cross-sectional region illuminated by structured light on the outer side of the wheel and rail; For the acquired image sequence of the structured light illumination section area on the outer side of the wheel and rail, the wheel and rail section profile point cloud sequence is reconstructed based on the stereo vision measurement principle and calibration parameters. Based on the reconstructed wheel-rail cross-section profile point cloud sequence, the plane is fitted in sections, and the angle of the normal vector of a specific plane is calculated and decomposed by combining multi-frame information to obtain the angle parameters. Based on the reconstructed wheel-rail cross-sectional profile point cloud sequence, the wheel circle is fitted by combining inter-frame and geometric constraints to determine the key points of the wheel-rail region and the rail reference plane, and the wheel-rail displacement parameters are calculated. The calculation of angle parameters includes: defining a coordinate system, wherein the coordinate system's... The axial direction is perpendicular to the extension direction of the rail and points outwards. The axial direction is parallel to the direction of rail extension. The axis is perpendicular to the ground and upwards; based on Real-time point cloud data Partition fitting along For the three planes along the axis from top to bottom, calculate the angle between the normal vectors of adjacent planes, and combine this with information from the previous frame to calculate the corrected angles. The rotation angle around the z-axis is the head-up angle; the rotation angle around the y-axis is the roll angle. Specifically, decompose the corrected angle between the normal vectors to obtain their respective head-up angle and roll angle; based on the decomposed head-up angle and roll angle, calculate... Always keep the head-shake angle and side roll angle in mind; Calculating wheel-rail displacement parameters includes: defining a coordinate system, where the coordinate system's... The axial direction is perpendicular to the extension direction of the rail and points outwards. The axial direction is parallel to the direction of rail extension. The axis is perpendicular to the ground and upwards; based on Real-time point cloud data Partition fitting along Three planes are used from top to bottom along the axis. Based on the normal vectors of the three planes, the equation of plane P is calculated. A corner detection algorithm is used to find feature points in the wheel area. The spatial circle containing the feature points is fitted by combining multi-frame motion and geometric constraints. The geometric constraints are: the center of the wheel circle has a finite range of motion in the z-axis direction; the radius of the wheel circle has geometric prior information. A fitting circle is defined, and the point on the fitting circle that is closest to plane P is considered as the key point in the wheel area. The displacement in the z-axis direction is the buoyancy; the displacement in the x-axis direction is the lateral displacement. The buoyancy is obtained by calculating the Euclidean distance between the key point in the wheel area in the z-axis direction and plane P, and the lateral displacement is obtained by calculating the Euclidean distance between the key point in the wheel area in the x-axis direction and plane P. Correct the angle between the two normal vectors and The equation is: ； in, , and The following are the fitting errors for the three planes, in order. , This indicates the confidence level of the included angle in the current frame. , This represents the mean confidence level of the first t-1 frames. , This represents the average angle of the first t-1 frames; calculate The corner of the head at all times and roll angle for: ； in , , and All are weighting coefficients; The equation for plane P is: ； in, , , These are the normal vectors of the three planes. For the first The spatial coordinates of the intersection points.

2. The non-contact online visual measurement method for wheel-rail dynamic attitude parameters according to claim 1, characterized in that, Acquiring an image sequence of the structured light irradiated section area outside the wheel and rail includes: calibrating the camera and structured light sensor using a standard checkerboard calibration board before acquiring the structured light images of the wheel and rail; triggering acquisition when the train passes to acquire an image sequence of the structured light irradiated section area outside the wheel and rail within the camera's field of view.

3. The non-contact online visual measurement method for wheel-rail dynamic attitude parameters according to claim 2, characterized in that, The wheel-rail cross-sectional profile point cloud sequence was reconstructed based on the parameters calibrated using a standard checkerboard calibration plate for the camera and structured light sensor.

4. A non-contact online visual measurement system for wheel-rail dynamic attitude parameters, characterized in that, include: The acquisition module is used to acquire image sequences of the structured light illumination section area outside the wheel and rail. The module is used to reconstruct the wheel-rail cross-sectional profile point cloud sequence based on the principle of stereo vision measurement and calibration parameters for the acquired image sequence of the structured light illumination section area on the outer side of the wheel and rail. The first calculation module is used to partition and fit planes based on the reconstructed wheel-rail profile point cloud sequence, and calculate and decompose the angle of the normal vector of a specific plane by combining multi-frame information to obtain the angle parameters. The second calculation module is used to fit the wheel circle based on the reconstructed wheel-rail cross-section profile point cloud sequence, combined with inter-frame and geometric constraints, to determine the key points of the wheel-rail region and the rail reference plane, and to calculate the wheel-rail displacement parameters. The calculation of angle parameters includes: defining a coordinate system, wherein the coordinate system's... The axial direction is perpendicular to the extension direction of the rail and points outwards. The axial direction is parallel to the direction of rail extension. The axis is perpendicular to the ground and upwards; based on Real-time point cloud data Partition fitting along For the three planes along the axis from top to bottom, calculate the angle between the normal vectors of adjacent planes, and combine this with information from the previous frame to calculate the corrected angles. The rotation angle around the z-axis is the head-up angle; the rotation angle around the y-axis is the roll angle. Specifically, decompose the corrected angle between the normal vectors to obtain their respective head-up angle and roll angle; based on the decomposed head-up angle and roll angle, calculate... Always keep the head-shake angle and side roll angle in mind; Calculating wheel-rail displacement parameters includes: defining a coordinate system, where the coordinate system's... The axial direction is perpendicular to the extension direction of the rail and points outwards. The axial direction is parallel to the direction of rail extension. The axis is perpendicular to the ground and upwards; based on Real-time point cloud data Partition fitting along Three planes are used from top to bottom along the axis. Based on the normal vectors of the three planes, the equation of plane P is calculated. A corner detection algorithm is used to find feature points in the wheel area. The spatial circle containing the feature points is fitted by combining multi-frame motion and geometric constraints. The geometric constraints are: the center of the wheel circle has a finite range of motion in the z-axis direction; the radius of the wheel circle has geometric prior information. A fitting circle is defined, and the point on the fitting circle that is closest to plane P is considered as the key point in the wheel area. The displacement in the z-axis direction is the buoyancy; the displacement in the x-axis direction is the lateral displacement. The buoyancy is obtained by calculating the Euclidean distance between the key point in the wheel area in the z-axis direction and plane P, and the lateral displacement is obtained by calculating the Euclidean distance between the key point in the wheel area in the x-axis direction and plane P. Correct the angle between the two normal vectors and The equation is: ； in, , and The following are the fitting errors for the three planes, in order. , This indicates the confidence level of the included angle in the current frame. , This represents the mean confidence level of the first t-1 frames. , This represents the average angle of the first t-1 frames; calculate The corner of the head at all times and roll angle for: ； in, , , and All are weighting coefficients; The equation for plane P is: ； in, , , These are the normal vectors of the three planes. For the first The spatial coordinates of the intersection points.

5. A non-transitory computer-readable storage medium, characterized in that, The non-transitory computer-readable storage medium is used to store computer instructions, which, when executed by a processor, implement the non-contact online visual measurement method for wheel-rail dynamic attitude parameters as described in any one of claims 1-3.

6. A computer program product, characterized in that, It includes a computer program, which, when run on one or more processors, is used to implement the non-contact online visual measurement method for wheel-rail dynamic attitude parameters as described in any one of claims 1-3.

7. An electronic device, characterized in that, include: The device includes a processor, a memory, and a computer program; wherein the processor is connected to the memory, and the computer program is stored in the memory. When the electronic device is running, the processor executes the computer program stored in the memory to cause the electronic device to execute instructions for implementing the online visual measurement method for non-contact wheel-rail dynamic attitude parameters as described in any one of claims 1-3.

Images