Three-dimensional eddy current thermal imaging detection device and method for fatigue cracks on surface of steel rail

Abstract

The invention discloses a three-dimensional eddy current thermal imaging detection device and method for fatigue cracks on the surface of a steel rail, and belongs to the field of material science, and the device comprises hardware equipment which comprises a detection motion platform, a mechanical arm device, a central control box, a three-dimensional eddy current thermal imaging module and an induction heating module; the mechanical arm device and the central control box are mounted at the top end of the detection motion platform, and the three-dimensional eddy current thermal imaging module and the induction heating module are both mounted at the output end of the mechanical arm device; the induction heating module is used for carrying out induction heating on the surface of a steel rail test block; the three-dimensional eddy current thermal imaging module is used for simultaneously acquiring infrared thermal image information and structured light three-dimensional information of the surface of the steel rail; and the software analysis system runs in the data acquisition and analysis module and is used for processing infrared thermal image information and structured light three-dimensional information through a three-dimensional eddy current thermal imaging crack detection algorithm and outputting a detection result. Infrared and structured light imaging results are fused, the detection precision is high, and the method is suitable for steel rail early-stage fatigue defect detection.

Description

Technical Field

[0001] This invention belongs to the field of materials science, and in particular relates to a three-dimensional eddy current thermal imaging detection device and method for fatigue cracks on the surface of steel rails. Background Technology

[0002] Fatigue cracks on rail surfaces are among the most common and somewhat concealed structural defects during rail service. If these cracks are not detected and quantitatively analyzed in their early stages, they will rapidly propagate and lead to serious accidents such as rail fracture, posing a significant threat to the safety of rail transit operations. Early-stage rail fatigue cracks often exhibit a closed morphology, making them difficult to identify effectively using conventional visual inspection, ultrasonic testing, and eddy current testing. These methods have limitations due to low signal-to-noise ratios and indistinct defect characteristics.

[0003] Addressing the challenges of inaccurate and quantifiable detection of fatigue cracks on rail surfaces, achieving highly sensitive, three-dimensional visualization detection of these cracks while overcoming the limitations of traditional methods in shallow crack identification, crack depth assessment, and environmental interference suppression is a critical technical challenge urgently needing to be solved in the field of nondestructive testing for rail tracks. To address this issue, there is a pressing need to develop a three-dimensional eddy current thermal imaging detection method and device that integrates eddy current thermal imaging and structured light features. This would enable high-precision cross-modal registration between visible light RGB images and thermal infrared images, overcoming challenges such as spectral appearance differences, low infrared resolution / contrast, and mismatches in scale, field of view, and viewing angle. Ultimately, this would improve the accuracy and robustness of fatigue crack detection on rail surfaces, providing reliable technical support for railway maintenance and track safety.

[0004] For the detection of fatigue cracks on rail surfaces, non-destructive testing methods such as eddy current testing, magnetic particle testing, visual inspection, and infrared thermal imaging are mainly used both domestically and internationally. These methods have achieved the identification of surface and near-surface defects of rails to a certain extent, but they still have problems such as limited sensitivity, insufficient quantitative ability, and poor anti-interference performance, making it difficult to meet the requirements for rapid and accurate identification of early cracks in complex service environments.

[0005] (1) Magnetic particle testing method

[0006] Magnetic particle testing, based on the principle of magnetic field distortion, can visually display the location of cracks and has high sensitivity to ferromagnetic materials. However, this method is only applicable to open cracks exposed on the surface, and its response to early fatigue cracks, closed cracks, or defects covered by oxides or contaminants is not obvious. In addition, the testing process requires surface cleaning and cumbersome magnetization and demagnetization, making it difficult to adapt to large-scale rapid testing in railway sites.

[0007] (2) Visual inspection method

[0008] Visual inspection technology can achieve high-resolution imaging detection of cracks on rail surfaces, making it suitable for identifying obvious surface damage or open cracks. However, its detection depends on lighting conditions and surface reflection characteristics, and its response to early cracks or subsurface defects is weak. Furthermore, the algorithm is easily affected by dirt and shadows, resulting in a certain probability of false detections.

[0009] (3) Eddy current detection method

[0010] Eddy current testing is widely used in rail inspection due to its non-contact nature and applicability to conductive materials. It identifies surface defects by detecting changes in electromagnetic induction signals, exhibiting high sensitivity. However, traditional eddy current testing methods are mostly single-frequency or low-dimensional, limiting their quantitative characterization of crack depth and spatial distribution. Furthermore, the detection signal is susceptible to variations in rail material, probe orientation, and the testing environment on the material surface, leading to fluctuations in the results.

[0011] (4) Infrared thermal imaging detection method

[0012] Infrared thermal imaging identifies the crack region by characterizing the temperature difference between the base material and the crack. This method offers advantages such as non-contact and rapid full-field imaging. However, thermal imaging methods are affected by ambient temperature fluctuations, differences in surface emissivity, and the thermal conductivity of materials, making it difficult to achieve high-precision quantitative detection.

[0013] In summary, existing technologies for detecting fatigue cracks on rail surfaces generally suffer from the following two problems: ① Insufficient ability to identify early, micro, or closed cracks, making it difficult to achieve early warning of cracks; ② Lack of three-dimensional spatial quantitative assessment methods, making it impossible to accurately assess crack depth and propagation trend.

[0014] Combining the advantages of eddy current and thermal imaging detection, eddy current thermal imaging can record the temperature distribution of surface / near-surface cracks in metallic materials and can detect fatigue cracks in rails. However, it suffers from problems such as non-uniform heating and field-of-view obstruction by the excitation coil. Therefore, there is an urgent need for those skilled in the art to propose a three-dimensional eddy current thermal imaging detection device and method for rail surface fatigue cracks, so as to achieve high-sensitivity identification and three-dimensional visualization quantitative assessment of rail surface fatigue cracks, and provide more reliable technical support for track safety monitoring and intelligent maintenance. Summary of the Invention

[0015] In view of this, the present invention provides a three-dimensional eddy current thermal imaging detection device and method for fatigue cracks on the surface of rails, to solve the above problems.

[0016] To achieve the above objectives, the present invention adopts the following technical solution: A three-dimensional eddy current thermal imaging detection device for fatigue cracks on the surface of rails includes: hardware equipment and software analysis system; The hardware device includes a motion detection platform, a robotic arm device, a central control box, a three-dimensional eddy current thermal imaging module, and an induction heating module; The robotic arm and the central control box are mounted on the top of the detection motion platform. The three-dimensional eddy current thermal imaging module and the induction heating module are both mounted on the output end of the robotic arm. The central control box is electrically connected to the detection motion platform, the robotic arm, the three-dimensional eddy current thermal imaging module, and the induction heating module, and is used for power supply, parameter setting, and data acquisition and processing. The induction heating module is used to induction heat the surface of the rail test block, and the three-dimensional eddy current thermal imaging module is used to simultaneously acquire infrared thermal image information and structured light three-dimensional information of the rail surface. The central control box is equipped with a data acquisition and analysis module. The software analysis system runs in the data acquisition and analysis module and is used to process the infrared thermal image information and structured light three-dimensional information through the three-dimensional eddy current thermal imaging crack detection algorithm, and output the detection results.

[0017] Furthermore, the detection motion platform includes a base for the detection device's traveling section and a power wheel; the power wheel is connected to the front and rear ends of the base for the detection device's traveling section and provides power to move on the surface of the rail test block.

[0018] Furthermore, the detection motion platform also includes auxiliary wheels, which are installed on both sides of the base of the traveling part of the detection device and roll on the flange of the rail test block; the auxiliary wheels have built-in displacement sensors to provide feedback on the displacement information of the detection device in order to locate defects.

[0019] Furthermore, the induction heating module includes an air-cooled induction heating device and a detection coil; the air-cooled induction heating device is electrically connected to the central control box and is used to receive excitation signals and drive the detection coil; during the detection process, the detection coil remains in a non-contact state with the surface of the rail test block, and the lifting distance is 2-4mm.

[0020] Furthermore, the three-dimensional eddy current thermal imaging module integrates an industrial camera, a projector, and an infrared thermal imager; the projector is used to project grating information onto the rail surface; the industrial camera is used to record the grating information projected onto the rail surface and generate visible light RGB images and depth point cloud data; the infrared thermal imager is used to acquire infrared thermal image sequences of the rail surface.

[0021] Furthermore, the central control box is also equipped with a power supply module and a test module; the power supply module supplies power to the entire device, and the test module is used to set test parameters and simultaneously trigger structured light acquisition and eddy current thermal imaging acquisition.

[0022] A three-dimensional eddy current thermal imaging method for detecting fatigue cracks on the surface of steel rails includes the following steps: S1: The surface of the rail test block is heated non-contactly using an induction heating module to generate an eddy current thermal field on the surface of the rail test block. S2: The infrared thermal image data and visible light RGB image data with structured light projected onto the surface of the rail test block are simultaneously acquired through the three-dimensional eddy current thermal imaging module. S3: Reconstruct the optical 3D point cloud image of the rail surface based on visible light RGB image data; S4: Employ a three-dimensional eddy current thermal imaging crack detection algorithm to perform cross-modal registration and fusion of infrared thermal image data and optical three-dimensional point cloud images; S5: Based on the fused data, extract the contour and depth information of cracks on the rail surface, generate a three-dimensional point set with temperature annotations, and realize crack detection.

[0023] Furthermore, in step S1, the diffusion equation of the induction heating source on the surface of the rail test block is expressed as:

[0024] in, It is the density of the rail material. It is the specific heat capacity of the rail material. It is the heat transfer coefficient. It's the detection speed. It is the temperature field distribution function. It is the electrical conductivity of the rail. It is the eddy current density on the surface of the rail.

[0025] Furthermore, in step S4, cross-modal registration includes a coarse registration stage and a fine registration stage: In the coarse registration stage: calculate the phase consistency amplitude and direction of the infrared thermal image and the structured light image; using the infrared thermal image as a reference, construct a multi-scale phase consistency histogram pyramid for global search to locate the region of maximum similarity between the structured light and the infrared thermal image; In the fine registration stage: second-order moment analysis is performed on the phase consistency map to detect significant structural points with modal invariance; phase consistency channel feature descriptors are constructed based on key points of the infrared image, the projection transformation type is estimated using the RANSAC algorithm, and the sub-pixel infrared image is resampled to the RGB coordinate system.

[0026] Further, step S5 specifically includes: extracting the surface crack features of the rail specimen from the infrared thermal image data by combining dilution matrix decomposition and skewness method; fusing the infrared features with the structured light point cloud information of the rail surface to obtain a likelihood image; accurately projecting the infrared thermal image onto the structured light three-dimensional point cloud to generate the three-dimensional thermal field result of the rail, thereby performing a three-dimensional visualization and quantitative assessment of fatigue cracks.

[0027] The beneficial effects of this invention are as follows: This invention achieves highly sensitive identification and three-dimensional quantitative assessment of fatigue cracks on rail surfaces by integrating infrared thermal imaging and structured light three-dimensional imaging technologies. The invention employs a detection coil made of flexible thermally conductive material, matched to the rail surface contour to ensure uniform heating. A two-stage registration algorithm, combining coarse and fine registration, effectively solves the problems of cross-modal image registration and fusion. Compared to traditional detection methods, this invention can not only accurately identify early-stage micro-closed cracks but also precisely quantify crack depth and spatial distribution, overcoming the influence of environmental interference and surface condition. Attached Figure Description

[0028] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0029] Figure 1 This is a schematic diagram of a three-dimensional eddy current thermal imaging detection device for fatigue cracks on the surface of steel rails.

[0030] Figure 2 This is a schematic diagram of the structure of a three-dimensional eddy current thermal imaging module.

[0031] Figure 3 This is a flowchart of a three-dimensional eddy current thermal imaging crack detection algorithm.

[0032] In the figure: 1-Base of the traveling part of the detection device, 2-Power wheel, 3-Auxiliary wheel, 4-Robotic arm device, 5-Central control box, 6-Three-dimensional eddy current thermal imaging module, 7-Air-cooled induction heating device, 8-Detection coil, 9-Rail test block, 10-Industrial camera, 11-Projector, 12-Infrared thermal imager. Detailed Implementation

[0033] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0034] Example 1

[0035] See attached document Figure 1-2A three-dimensional eddy current thermal imaging detection device for fatigue cracks on rail surfaces includes: hardware equipment and software analysis system; the hardware equipment includes a detection motion platform, a robotic arm device 4, a central control box 5, a three-dimensional eddy current thermal imaging module 6, and an induction heating module; the robotic arm device 4 and the central control box 5 are installed on the top of the detection motion platform, and the three-dimensional eddy current thermal imaging module 6 and the induction heating module are both installed on the output end of the robotic arm device 4, wherein the robotic arm device 4 is a six-axis robotic arm, and the robotic arm device 4 can flexibly adjust the testing angle according to the type of rail being tested; the central control box 5 is connected to the detection platform and the central control box 5. The motion platform, robotic arm device 4, three-dimensional eddy current thermal imaging module 6, and induction heating module are electrically connected for power supply, parameter setting, and data acquisition and processing. The induction heating module is used to induction heat the surface of the rail test block 9, and the three-dimensional eddy current thermal imaging module 6 is used to simultaneously acquire infrared thermal image information and structured light three-dimensional information of the rail surface. The central control box 5 is equipped with a data acquisition and analysis module, and the software analysis system runs in the data acquisition and analysis module to process the infrared thermal image information and structured light three-dimensional information through the three-dimensional eddy current thermal imaging crack detection algorithm and output the detection results.

[0036] In a preferred embodiment, the detection motion platform includes a detection device travel section base 1 and a power wheel 2; the power wheel 2 is connected to the front and rear ends of the detection device travel section base 1 and provides power to move on the surface of the rail test block 9.

[0037] In a preferred embodiment, the detection motion platform further includes auxiliary wheels 3, which are installed on both sides of the base 1 of the traveling part of the detection device and roll on the flange of the rail test block 9. The auxiliary wheels 3 have built-in displacement sensors to provide feedback on the displacement information of the detection device in order to locate defects.

[0038] In a preferred embodiment, the induction heating module includes an air-cooled induction heating device 7 and a detection coil 8. The air-cooled induction heating device 7 is electrically connected to the central control box 5 and is used to receive excitation signals and drive the detection coil 8. During the detection process, the air-cooled induction heating device 7 and the detection coil 8 work together to generate temperature on the surface of the rail test block 9. In order to balance heating efficiency and non-contact testing, the detection coil 8 should maintain a lifting distance of 2-4 mm from the surface of the rail test block 9.

[0039] In a preferred embodiment, the three-dimensional eddy current thermal imaging module 6 integrates an industrial camera 10, a projector 11, and an infrared thermal imager 12; the projector 11 is used to project grating information onto the rail surface; the industrial camera 10 is used to record the grating information projected onto the rail surface and generate visible light RGB images and depth point cloud data; the infrared thermal imager 12 is used to acquire infrared thermal image sequences of the rail surface.

[0040] In a preferred embodiment, the central control box 5 is further provided with a power supply module and a test module; the power supply module supplies power to the entire device, and the test module is used to set test parameters and synchronously trigger structured light acquisition and eddy current thermal imaging acquisition.

[0041] Example 2

[0042] See attached document Figure 3 A three-dimensional eddy current thermal imaging detection method for fatigue cracks on rail surfaces includes the following steps: S1: The surface of the rail test block 9 is heated non-contactly using an induction heating module to generate an eddy current thermal field on the surface of the rail test block 9. The specific implementation plan includes both static and dynamic testing types, both with the same heat generation mechanism. When the central control box 5 outputs a high-frequency AC signal, the air-cooled induction heating device 7 amplifies the signal and transmits it to the detection coil 8. The detection coil 8 is not in contact with the rail test block 9. During the propagation of the electromagnetic wave in space, it carries energy and forms an eddy current signal with the same frequency as the excitation signal on the surface of the rail test block 9. This energy is converted into Joule heat through eddy current loss on the material surface, and this dissipated power forms the induction heating source. Regarding heat conduction, the heat conduction theory in the static testing process can provide the temperature distribution law, but it is not applicable to the dynamic testing theory. Because in the dynamic testing theory, there is relative motion between the heating source and the test piece, the temperature distribution on the material surface will vary depending on the testing speed. Therefore, the diffusion equation of the induction heating source on the surface of the rail test block 9 can be expressed as:

[0043] in, It is the density of the rail material. It is the specific heat capacity of the rail material. It is the heat transfer coefficient. It's the detection speed. It is the temperature field distribution function. It is the electrical conductivity of the rail. It is the eddy current density on the surface of the rail.

[0044] S2: The infrared thermal image data and visible light RGB image data with structured light projected on the surface of the rail test block 9 are simultaneously acquired by the three-dimensional eddy current thermal imaging module 6. S3: Reconstruct the optical 3D point cloud image of the rail surface based on visible light RGB image data; S4: Employ a three-dimensional eddy current thermal imaging crack detection algorithm to perform cross-modal registration and fusion of infrared thermal image data and optical three-dimensional point cloud images; Cross-modal registration includes a coarse registration stage and a fine registration stage: In the coarse registration stage: calculate the phase consistency amplitude and direction of the infrared thermal image and the structured light image; using the infrared thermal image as a reference, construct a multi-scale phase consistency histogram pyramid for global search to locate the region of maximum similarity between the structured light and the infrared thermal image; In the fine registration stage: second-order moment analysis is performed on the phase consistency map to detect significant structural points with mode invariance; phase consistency channel feature descriptors are constructed based on key points of the infrared image, the projection transformation type is estimated using the RANSAC algorithm, and the infrared image is subpixel resampled to the RGB coordinate system; S5: Based on the fused data, the contour and depth information of cracks on the rail surface are extracted to generate a three-dimensional point set with temperature annotations, enabling crack detection. Specifically, this includes: extracting surface crack features of rail specimen 9 from infrared thermal image data by combining dilution matrix decomposition and skewness methods; fusing the infrared features with the structured light point cloud information of the rail surface to obtain a likelihood image; and accurately projecting the infrared thermal image onto the structured light three-dimensional point cloud to generate a three-dimensional thermal field result for the rail, thereby enabling three-dimensional visualization and quantitative assessment of fatigue cracks.

[0045] The above descriptions are merely specific embodiments of the present invention, and common knowledge regarding the specific structures and characteristics of the solutions is not described in detail here. It should be noted that those skilled in the art can make various modifications and improvements without departing from the structure of the present invention, and these should also be considered within the scope of protection of the present invention. These modifications and improvements will not affect the effectiveness of the implementation of the present invention or the practicality of the patent. The scope of protection claimed in this application should be determined by the content of its claims, and the specific embodiments described in the specification can be used to interpret the content of the claims.

[0046] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to the method section.

[0047] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A three-dimensional eddy current thermal imaging detection device for fatigue cracks on the surface of steel rails, characterized in that, include: Hardware equipment and software analysis systems; The hardware device includes a motion detection platform, a robotic arm device (4), a central control box (5), a three-dimensional eddy current thermal imaging module (6), and an induction heating module; The robotic arm device (4) and the central control box (5) are installed on the top of the detection motion platform. The three-dimensional eddy current thermal imaging module (6) and the induction heating module are both installed at the output end of the robotic arm device (4). The central control box (5) is electrically connected to the detection motion platform, the robotic arm device (4), the three-dimensional eddy current thermal imaging module (6) and the induction heating module respectively, and is used for power supply, parameter setting and data acquisition and processing. The induction heating module is used to induction heat the surface of the rail test block (9), and the three-dimensional eddy current thermal imaging module (6) is used to simultaneously acquire infrared thermal image information and structured light three-dimensional information of the rail surface. The central control box (5) is equipped with a data acquisition and analysis module. The software analysis system runs in the data acquisition and analysis module and is used to process the infrared thermal image information and structured light three-dimensional information through the three-dimensional eddy current thermal imaging crack detection algorithm and output the detection results.

2. The three-dimensional eddy current thermal imaging detection device for fatigue cracks on the surface of rails according to claim 1, characterized in that, The detection motion platform includes a detection device running base (1) and a power wheel (2); the power wheel (2) is connected to the front and rear ends of the detection device running base (1) and provides power to move on the surface of the rail test block (9).

3. The three-dimensional eddy current thermal imaging detection device for fatigue cracks on the surface of rails according to claim 2, characterized in that, The detection motion platform also includes auxiliary wheels (3), which are installed on both sides of the base (1) of the traveling part of the detection device and roll on the flange of the rail test block (9); the auxiliary wheels (3) are equipped with displacement sensors to provide feedback on the displacement information of the detection device in order to locate defects.

4. The three-dimensional eddy current thermal imaging detection device for fatigue cracks on the surface of rails according to claim 1, characterized in that, The induction heating module includes an air-cooled induction heating device (7) and a detection coil (8); the air-cooled induction heating device (7) is electrically connected to the central control box (5) and is used to receive excitation signals and drive the detection coil (8); during the detection process, the detection coil (8) remains in a non-contact state with the surface of the rail test block (9) and is lifted away by a distance of 2-4 mm.

5. The three-dimensional eddy current thermal imaging detection device for fatigue cracks on the surface of rails according to claim 1, characterized in that, The three-dimensional eddy current thermal imaging module (6) integrates an industrial camera (10), a projector (11), and an infrared thermal imager (12); the projector (11) is used to project grating information onto the surface of the rail; the industrial camera (10) is used to record the grating information projected onto the surface of the rail and generate visible light RGB images and depth point cloud data; the infrared thermal imager (12) is used to collect infrared thermal image sequences of the surface of the rail.

6. The three-dimensional eddy current thermal imaging detection device for fatigue cracks on the surface of rails according to claim 1, characterized in that, The central control box (5) is also equipped with a power supply module and a test module; the power supply module supplies power to the entire device, and the test module is used to set test parameters and synchronously trigger structured light acquisition and eddy current thermal imaging acquisition.

7. A three-dimensional eddy current thermal imaging detection method for fatigue cracks on the surface of steel rails, characterized in that, Includes the following steps: S1: The surface of the rail test block (9) is heated in a non-contact manner using an induction heating module, so that an eddy current thermal field is generated on the surface of the rail test block (9). S2: The infrared thermal image data and visible light RGB image data with structured light projected on the surface of the rail test block (9) are simultaneously acquired by the three-dimensional eddy current thermal imaging module (6); S3: Reconstruct the optical 3D point cloud image of the rail surface based on visible light RGB image data; S4: Employ a three-dimensional eddy current thermal imaging crack detection algorithm to perform cross-modal registration and fusion of infrared thermal image data and optical three-dimensional point cloud images; S5: Based on the fused data, extract the contour and depth information of cracks on the rail surface, generate a three-dimensional point set with temperature annotations, and realize crack detection.

8. The three-dimensional eddy current thermal imaging detection method for fatigue cracks on the surface of rails according to claim 7, characterized in that, In step S1, the diffusion equation of the induction heating source on the surface of the rail test block (9) is expressed as: in, It is the density of the rail material. It is the specific heat capacity of the rail material. It is the heat transfer coefficient. It's the detection speed. It is the temperature field distribution function. It is the electrical conductivity of the rail. It is the eddy current density on the surface of the rail.

9. The three-dimensional eddy current thermal imaging detection method for fatigue cracks on the surface of rails according to claim 7, characterized in that, In step S4, cross-modal registration includes a coarse registration stage and a fine registration stage: In the coarse registration stage: calculate the phase consistency amplitude and direction between the infrared thermal image and the structured light image; Based on infrared thermal images, a multi-scale phase consistency histogram pyramid is constructed for global search to locate the region of maximum similarity between structured light and infrared thermal images. In the fine registration stage: Second-order moment analysis is performed on the phase consistency map to detect significant structural points with modal invariance; Phase-consistent channel feature descriptors are constructed based on key points of infrared images. The RANSAC algorithm is used to estimate the projection transformation type, and the sub-pixel infrared images are resampled to the RGB coordinate system.

10. A three-dimensional eddy current thermal imaging detection method for fatigue cracks on the surface of a rail according to claim 7, characterized in that, Step S5 specifically includes: combining dilution matrix decomposition and skewness method to extract the surface crack features of rail test block (9) from infrared thermal image data; fusing infrared features with the structured light point cloud information of rail surface to obtain a likelihood image; accurately projecting the infrared thermal image onto the structured light three-dimensional point cloud to generate the three-dimensional thermal field result of rail, thereby performing a three-dimensional visualization quantitative assessment of fatigue cracks.

Images