A method and apparatus for rail bond testing

Abstract

This invention provides a method and apparatus for detecting rail grounding. It acquires a third rail image and a first set of rail images, processes the first set of rail images to obtain a second rail image, and then compares the pixels in the second and third rail images to obtain a fourth rail image. The fourth rail image is the second rail image whose pixel grayscale value is greater than a preset grayscale threshold compared to the third rail image, indicating an anomaly. Based on the fourth rail image, a fifth rail image is acquired. A first value list and a second value list are obtained based on the pixel values ​​of the fifth rail image, thereby determining whether the abnormal pixel grayscale value is caused by foreign objects or rust. Therefore, this invention can detect not only foreign objects on the rail but also rust.

Description

Technical Field

[0001] This invention relates to the field of railway track inspection technology, and in particular to a method and apparatus for railway track grounding detection. Background Technology

[0002] Rail transit is a major mode of passenger and freight transport in the country and has significant strategic importance. Among these, railway transport plays a particularly crucial role. Due to my country's large population and high-volume rail traffic, more rigorous security checks and inspections are necessary to ensure the safety of passengers and goods.

[0003] In the current technology, with the widespread application of low-altitude drones in the field of inspection and surveillance, more and more low-altitude drones are being used for railway line inspections to save manpower and material costs. Railway maintenance personnel use the videos or pictures of the railway tracks transmitted by low-altitude drones to inspect the condition of the tracks, and promptly carry out maintenance when foreign objects are found on the tracks.

[0004] In existing technologies, to ensure the safe operation of railways, images transmitted by drones are manually verified, which consumes a lot of manpower and resources. In addition, extracting multi-level depth features from images transmitted by drones to determine whether there are foreign objects on the rails consumes a lot of computing time. Furthermore, existing methods only detect foreign objects on the rails and do not detect the rust condition on the rails. Summary of the Invention

[0005] To address the aforementioned technical problems, the technical solution adopted by this invention is as follows:

[0006] A method for detecting railway rail grounding, the method comprising the following steps:

[0007] S100. Obtain the first set of railway track images A = {A1, ..., A2} taken by the aircraft. i , ..., A m}, A i ={A i1 , ..., A ij , ..., A in}, A ij Let j be the j-th rail image taken by the aircraft on day i, where j can be 1 to n, n is the number of rail images taken by the aircraft on day i, i can be 1 to m, and m is the number of images in the first rail image list.

[0008] S200. Process the first railway track image set to obtain the second railway track image set B = {B1, ..., B...} i , ..., B m}, B i ={B i1 , ..., B ij , ..., B in}, B ij = (B 0 ij RB 0x ij ), where B 0 ij Let RB be the ID of the j-th second rail image in the i-th second rail image list. 0x ij For B 0 ij The grayscale value corresponding to the x-th pixel in the image, where x ranges from 1 to q, and q is a value of B. 0 ij The number of corresponding pixels.

[0009] S300, Obtain the image of the third railway track C 0 ={C 0 1, ..., C 0 x , ..., C 0 q}, C 0 x Let x be the grayscale values ​​of x pixels in the third railway track image.

[0010] S400, according to C 0 Together with B, obtain the fourth rail image D. 0 D 0 ={D 0 1, ..., D 0 g , ..., D 0 z}, D 0 g =(XD 0 g YD 0 g RD 0 g ), XD 0 g D 0 The corresponding x-axis coordinate of the fourth pixel (g-th pixel), YD 0 g D 0 The corresponding y-axis coordinate of the fourth pixel of the g-th pixel, RD 0 g D 0 The corresponding grayscale value of the fourth pixel at the g-th position, where g ranges from 1 to z, and z is D. 0 The corresponding number of fourth pixels, where the fourth pixel is |RB 0x ij -C 0 x |>r0 The corresponding second pixel, r 0 This is the preset grayscale threshold.

[0011] S500. Based on D, obtain the fifth railway track image list E = {E1, ..., E...} y , ..., E p}, E y ={E y1 , ..., E yg , ..., E yz}, E yg Let y be the g-th fifth pixel in the y-th fifth rail image, where y ranges from 1 to p, and p is the number of fifth rail images. The coordinates of the fifth pixel are the same as those of the corresponding fourth pixel.

[0012] S600. According to E, obtain the first numerical list S1 = {S11, ..., S1...} g , ..., S1 z}, where S1 g The following conditions must be met:

[0013] S1 g =∑ p y＝2 (RE yg -RE (y-1)g ) / p-1;

[0014] Among them, RE yg For E yg The corresponding grayscale value.

[0015] S700, when S1 g <k 0 When k is in time, output the first prompt message, where k 0 This is a preset first numerical threshold.

[0016] S800, when S1 g ≥k 0 At that time, based on S1, obtain the second numerical list S2 = {S21, ..., S2}. e S2 h}, S2 e Let S2 be the e-th second value, where e ranges from 1 to h, and h is the number of second values. e The following conditions must be met:

[0017] S2 e =RD 0 e -RE ye ;

[0018] Among them, RD 0 e D0 The corresponding grayscale value of the fourth pixel, RE ye Let be the grayscale value of the e-th fifth pixel.

[0019] S900, when S2 e >k′ 0 When the time comes, output the first prompt message, k′ 0 This is a preset second numerical threshold.

[0020] S1000, when S2 e ≤k′ 0 And S2 e When the preset target conditions are met, a second prompt message is output.

[0021] The present invention has at least the following beneficial effects:

[0022] (1) By obtaining the first intermediate rail image and the number of key pixels, and processing the first intermediate rail image to unify the number of pixels in the target area of ​​the first intermediate rail image to the number of key pixels, the second intermediate rail image is obtained. Then, the second intermediate rail image is grayscaled to obtain the second rail image. This unifies the number of pixels in the first rail image and solves the problem of inconsistent resolution of the first rail image caused by the altitude of the aircraft. Therefore, subsequent processing is more convenient and time efficiency is improved.

[0023] (2) By acquiring the third rail image and the first rail image set, and processing the first rail image set to acquire the second rail image, the second rail image is then compared with the pixels in the third rail image to acquire the fourth rail image. The fourth rail image is the second rail image whose pixel gray value is greater than the gray value of the third rail image than a preset gray value threshold, which is the second rail image with an abnormality. Based on the fourth rail image, the fifth rail image is acquired. Based on the pixel value of the fifth rail image, the first value list and the second value list are acquired to determine whether the abnormal pixel gray value is caused by foreign objects or rust on the rail. Therefore, the present invention can not only detect the presence of foreign objects on the rail, but also detect the rust on the rail. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying 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.

[0025] Figure 1 A schematic flowchart of a method for detecting railway grounding provided in an embodiment of the present invention;

[0026] Figure 2 This is a schematic diagram of the device for detecting railway grounding provided in an embodiment of the present invention. Detailed Implementation

[0027] 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.

[0028] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or server that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0029] This invention provides a method for detecting railway grounding, such as... Figure 1 As shown, the method includes the following steps:

[0030] S100. Obtain the first set of railway track images A = {A1, ..., A2} taken by the aircraft. i , ..., A m}, A i ={A i1 , ..., A ij , ..., A in}, A ij Let j be the j-th rail image taken by the aircraft on day i, where j can be 1 to n, n is the number of rail images taken by the aircraft on day i, i can be 1 to m, and m is the number of images in the first rail image list.

[0031] Preferably, the aircraft is a drone.

[0032] Furthermore, n satisfies the following condition:

[0033] n = 24 * 60 * 60 * k 0 ;

[0034] Where, k0 Preset shooting threshold;

[0035] Furthermore, those skilled in the art can set k according to actual needs. 0 This will not be elaborated upon here.

[0036] S200. Process the first railway track image set to obtain the second railway track image set B = {B1, ..., B...} i , ..., B m}, B i ={B i1 , ..., B ij , ..., B in}, B ij = (B 0 ij RB 0x ij ), where B 0 ij Let RB be the ID of the j-th second rail image in the i-th second rail image list. 0x ij For B 0 ij The grayscale value corresponding to the x-th pixel in the image, where x ranges from 1 to q, and q is a value of B. 0 ij The number of corresponding pixels.

[0037] Specifically, the second rail image ID is a unique identifier for the second rail image.

[0038] Furthermore, S200 also includes the following steps to obtain a second set of railway track images:

[0039] S210. Based on A, obtain the first intermediate rail image set F = {F1, ..., F2}. i , ..., F m}, F i ={F i1 , ..., F ij , ..., F in}, F ij =(F 0 ij SF 0 ij ), where F 0 ij For the i-th first intermediate rail image list, the j-th first intermediate rail image ID is SF. 0 ij For F 0 ij The number of pixels in the corresponding target area.

[0040] Specifically, those skilled in the art will know that any method for traversing the first intermediate rail image to obtain the number of pixels corresponding to the first intermediate rail image falls within the protection scope of this invention, and will not be elaborated here.

[0041] Furthermore, in this embodiment of the invention, the target area is the area from the leftmost side of the rail to the rightmost side of the rail when the train is traveling. It can be understood that the area on both sides of the rail and the middle sleeper area is the target area.

[0042] Furthermore, the first intermediate rail image is a first rail image that only contains the target area. It can be understood as a first rail image being cropped to extract the portion of the first rail image excluding the target area.

[0043] Furthermore, those skilled in the art will know that any method for cropping and processing the first railway track image to obtain the first intermediate railway track image falls within the protection scope of this invention, and will not be elaborated further here.

[0044] S230. Based on F, obtain the number of key pixels q, where q satisfies the following steps:

[0045] q=∑ m i＝1 (SG 0 i ) / m;

[0046] Among them, SG 0 i =∑ n j＝1 (SF 0 ij ) / n.

[0047] Specifically, the number of key pixels is the number of pixels corresponding to the third railway track image.

[0048] S250. Process F to obtain the second intermediate rail image set F′={F′1,……,F′ i , ..., F′ m}, F′ i ={F′ i1 , ..., F′ ij , ..., F′ in}, F′ ij The j-th second intermediate rail image is the i-th second intermediate rail image in the list of second intermediate rail images. The second intermediate rail image is the first intermediate rail image where the number of pixels in the target area is uniformly q.

[0049] Specifically, those skilled in the art will know that any method for unifying the second intermediate rail image to a fixed value falls within the scope of protection of this invention, such as Nearest Interpolation, Bilinear Interpolation, and Bi-cubic scaling.

[0050] Preferably, in this embodiment of the invention, bilinear interpolation is used to process F. The image obtained by bilinear interpolation is smoother than that obtained by nearest neighbor interpolation, and the algorithm is simpler than that of bicubic interpolation. Therefore, while improving time efficiency, the accuracy of the second intermediate rail image is guaranteed.

[0051] S270. Process F′ to obtain B.

[0052] Specifically, those skilled in the art will know that any method for grayscale processing of the second intermediate rail image falls within the protection scope of this invention, such as the maximum value method, the average value method, and the weighted average method.

[0053] Preferably, in this embodiment of the invention, the average value method is used to perform grayscale processing on F′. The average value is obtained by averaging the three brightness components in the color image to obtain a grayscale value. The calculation is small and the grayscale image is clean, thus improving time efficiency.

[0054] As described above, S210-S230 obtains the first intermediate rail image and the number of key pixels, processes the first intermediate rail image to unify the number of pixels in the target area of ​​the first intermediate rail image to the number of key pixels, obtains the second intermediate rail image, and then performs grayscale processing on the second intermediate rail image to obtain the second rail image. This unifies the number of pixels in the first rail image and solves the problem of inconsistent resolution of the first rail image caused by the altitude of the aircraft. Therefore, it makes subsequent processing more convenient and improves time efficiency.

[0055] S300, Obtain the image of the third railway track C 0 ={C 0 1, ..., C 0 x , ..., C 0 q}, C 0 x Let x be the grayscale values ​​of x pixels in the third railway track image.

[0056] Specifically, the third rail image is a rail image obtained by taking a picture of the rail in its optimal state, processing it to determine the number of key pixels, and then performing grayscale processing.

[0057] S400, according to C 0 Together with B, obtain the fourth rail image D. 0 D 0 ={D 0 1, ..., D 0 g , ..., D 0 z}, D 0 g =(XD 0 g YD 0 g RD 0 g ), XD 0 g D 0 The corresponding x-axis coordinate of the fourth pixel (g-th pixel), YD 0 g D 0 The corresponding y-axis coordinate of the fourth pixel of the g-th pixel, RD 0 g D 0 The corresponding grayscale value of the fourth pixel at the g-th position, where g ranges from 1 to z, and z is D. 0 The corresponding number of fourth pixels, where the fourth pixel is |RB 0x ij -C 0 x |>r 0 The corresponding second pixel, r 0 This is the preset grayscale threshold.

[0058] Specifically, those skilled in the art will know that any method for obtaining the horizontal axis coordinates and vertical axis coordinates of the fourth track image pixels falls within the protection scope of this invention, and will not be elaborated further here.

[0059] Furthermore, OpenCV was used to process the second rail image to obtain XD. 0 g and YD 0 g .

[0060] Furthermore, those skilled in the art can set r according to actual needs. 0 This will not be elaborated upon here.

[0061] S500. Based on D, obtain the fifth railway track image list E = {E1, ..., E...} y , ..., E p}, E y ={E y1, ..., E yg , ..., E yz}, E yg Let g be the fifth pixel in the y-th fifth rail image, where y ranges from 1 to p, and p is the number of fifth rail images. The coordinates of the fifth pixel are the same as those of the corresponding fourth pixel.

[0062] Specifically, E y ∈[t 0 -△t,t 0 This can be understood as the time difference between capturing the fifth rail image and the fourth rail image not exceeding Δt, where t 0 The time point at which the image of the fourth railway track was captured is Δt, which is a preset time threshold.

[0063] Furthermore, those skilled in the art can set Δt according to actual needs, which will not be elaborated here.

[0064] Furthermore, the S500 also includes the following steps:

[0065] S510, according to E y , obtain E y The corresponding x-axis coordinate XE of the g-th fifth pixel yg and E y The corresponding ordinate YE of the fifth pixel (g-th pixel) yg ;

[0066] S530, according to XE yg and YE yg Get RE yg .

[0067] S600. According to E, obtain the first numerical list S1 = {S11, ..., S1...} g , ..., S1 z}, where S1 g The following conditions must be met:

[0068] S1 g =∑ p y＝2 (RE yg -RE (y-1)g ) / p-1;

[0069] Among them, RE yg For E yg The corresponding grayscale value.

[0070] S700, when S1 g <k 0 When k is in time, output the first prompt message, where k 0 This is a preset first numerical threshold.

[0071] Specifically, those skilled in the art can set k according to actual needs. 0 This will not be elaborated upon here.

[0072] Furthermore, the first notification message indicates the presence of a foreign object system on the railway track.

[0073] S800, when S1 g ≥k 0 At that time, based on S1, obtain the second numerical list S2 = {S21, ..., S2}. e S2 h}, S2 e Let S2 be the e-th second value, where e ranges from 1 to h, and h is the number of second values. e The following conditions must be met:

[0074] S2 e =RD 0 e -RE ye ;

[0075] Among them, RD 0 e D 0 The corresponding grayscale value of the fourth pixel, RE ye Let be the grayscale value of the e-th fifth pixel.

[0076] S900, when S2 e >k′ 0 When the time comes, output the first prompt message, k′ 0 This is a preset second numerical threshold.

[0077] Specifically, those skilled in the art can set k′ according to actual needs. 0 This will not be elaborated upon here.

[0078] S1000, when S2 e ≤k′ 0 And S2 e When the preset target conditions are met, a second prompt message is output.

[0079] Specifically, the target condition is E 1g >...>E yg >...>E pg or E 1g ≤……E yg ≤……≤E pg .

[0080] Furthermore, the second notification message is a notification message indicating that the rails are rusty.

[0081] As described above, S100-S1000 acquires a third rail image and a first rail image set, processes the first rail image set to acquire a second rail image, compares the pixels in the second rail image with those in the third rail image to acquire a fourth rail image. The fourth rail image is the second rail image whose pixel grayscale value is greater than the grayscale value of the third rail image than a preset grayscale threshold, i.e., the second rail image with an anomaly. Based on the fourth rail image, a fifth rail image is acquired. Based on the pixel values ​​of the fifth rail image, a first value list and a second value list are acquired to determine whether the abnormal pixel grayscale value is caused by foreign objects or rust on the rail. Therefore, this invention can not only detect the presence of foreign objects on the rail, but also detect the rust on the rail.

[0082] like Figure 2 As shown, the device for detecting rail grounding provided in this embodiment of the invention includes: an acquisition module 1, an image processing module 2, an abnormal pixel extraction module 3, and a judgment module 4.

[0083] Specifically, the acquisition module 1 is used to acquire a first rail image, which is a rail image taken by an aircraft; the image processing module 2 is used to process the first rail image to acquire a second rail image; the abnormal pixel extraction module 3 is used to acquire a fourth rail image, which is a second rail image with abnormal pixel grayscale values; the judgment module 4 is used to determine whether the abnormal pixels in the fourth rail image are in a first abnormal state or a second abnormal state according to the fifth rail image list, where the first abnormal state is the presence of foreign objects on the rail and the second abnormal state is the rusting of the rail.

[0084] Embodiments of the present invention also provide a non-transitory computer-readable storage medium that can be disposed in an electronic device to store at least one instruction or at least one program related to implementing a method in the method embodiments, wherein the at least one instruction or the at least one program is loaded and executed by the processor to implement the method provided in the above embodiments.

[0085] Embodiments of the present invention also provide an electronic device, including a processor and the aforementioned non-transitory computer-readable storage medium.

[0086] While specific embodiments of the invention have been described in detail by way of example, those skilled in the art should understand that the above examples are for illustrative purposes only and are not intended to limit the scope of the invention. Those skilled in the art should also understand that various modifications can be made to the embodiments without departing from the scope and spirit of the invention. The scope of the invention is defined by the appended claims.

Claims

1. A method for detecting railway rail grounding, characterized in that, The method includes the following steps: S100. Obtain the first set of railway track images A = {A1, ..., A2} taken by the aircraft. i , ..., A m }, A i ={A i1 , ..., A ij , ..., A in }, A ij Let j be the j-th rail image taken by the aircraft on day i, where j can be 1 to n, n is the number of rail images taken by the aircraft on day i, i can be 1 to m, and m is the number of images in the first rail image list. S200. Process the first railway track image set to obtain the second railway track image set B = {B1, ..., B...} i , ..., B m }, B i ={B i1 , ..., B ij , ..., B in }, B ij = (B 0 ij RB 0x ij ), where B 0 ij Let RB be the ID of the j-th second rail image in the i-th second rail image list. 0x ij For B 0 ij The grayscale value corresponding to the x-th pixel in the image, where x ranges from 1 to q, and q is a value of B. 0 ij The corresponding number of pixels; S300, Obtain the image of the third railway track C 0 ={C 0 1, ..., C 0 x , ..., C 0 q }, C 0 x Let x be the grayscale values ​​of x pixels in the third railway track image; S400, according to C 0 Together with B, obtain the fourth rail image D. 0 D 0 ={D 0 1, ..., D 0 g , ..., D 0 z }, D 0 g =(XD 0 g YD 0 g RD 0 g ), XD 0 g D 0 The corresponding x-axis coordinate of the fourth pixel (g-th pixel), YD 0 g D 0 The corresponding y-axis coordinate of the fourth pixel of the g-th pixel, RD 0 g D 0 The corresponding grayscale value of the fourth pixel at the g-th position, where g ranges from 1 to z, and z is D. 0 The corresponding number of fourth pixels, where the fourth pixel is |RB 0x ij -C 0 x |>r 0 The corresponding second pixel, r 0 The preset grayscale value threshold is used; S500. Based on D, obtain the fifth railway track image list E = {E1, ..., E...} y , ..., E p }, E y ={E y1 , ..., E yg , ..., E yz }, E yg Let g be the fifth pixel in the y-th fifth rail image, where y ranges from 1 to p, and p is the number of fifth rail images. The coordinates of the fifth pixel are the same as those of the corresponding fourth pixel. S600. According to E, obtain the first numerical list S1 = {S11, ..., S1...} g , ..., S1 z }, where S1 g The following conditions must be met: S1 g =∑ p y＝2 (RE yg -RE (y-1)g ) / p-1; Among them, RE yg For E yg The corresponding grayscale value; S700, when S1 g <k 0 When k is in time, output the first prompt message, where k 0 The first numerical threshold is preset; S800, when S1 g ≥k 0 At that time, based on S1, obtain the second numerical list S2 = {S21, ..., S2}. e S2 h }, S2 e Let S2 be the e-th second value, where e ranges from 1 to h, and h is the number of second values. e The following conditions must be met: S2 e =RD 0 e -RE ye ; Among them, RD 0 e D 0 The corresponding grayscale value of the fourth pixel, RE ye Let be the grayscale value of the e-th fifth pixel. S900, when S2 e >k′ 0 When the time comes, output the first prompt message, k′ 0 A preset second numerical threshold; S1000, when S2 e ≤k′ 0 And S2 e When the preset target conditions are met, a second prompt message is output.

2. The method according to claim 1, characterized in that, S200 also includes the following steps: S210. Based on A, obtain the first intermediate rail image set F = {F1, ..., F2}. i , ..., F m }, F i ={F i1 , ..., F ij , ..., F in }, F ij =(F 0 ij SF 0 ij ), where F 0 ij For the i-th first intermediate rail image list, the j-th first intermediate rail image ID is SF. 0 ij For F 0 ij The number of pixels in the corresponding target area; S230. Based on F, obtain the number of key pixels q, where q satisfies the following steps: q=∑ m i＝1 (SG 0 i ) / m; Among them, SG 0 i =∑ n j＝1 (SF 0 ij ) / n; S250. Process F to obtain the second intermediate rail image set F′={F′1,……,F′ i , ..., F′ m }, F′ i ={F′ i1 , ..., F′ ij , ..., F′ in }, F′ ij The j-th second intermediate rail image is the i-th second intermediate rail image in the list of second intermediate rail images. The second intermediate rail image is the first intermediate rail image where the number of pixels in the target area is uniformly q. S270. Process F′ to obtain B.

3. The method according to claim 1, characterized in that, The third rail image is a rail image without any abnormalities.

4. The method according to claim 1, characterized in that, In the S400, OpenCV is used to process the image of the second railway track to obtain XD. 0 g and YD 0 g .

5. The method according to claim 1, characterized in that, In S700, the target condition is E. 1g >...>E yg >...>E pg or E 1g ≤……E yg ≤……≤E pg .

6. The method according to claim 2, characterized in that, In S210, the target area is the area from the leftmost side of the rail to the rightmost side of the rail when the train is traveling.

7. The method according to claim 2, characterized in that, In S250, bilinear interpolation is used to process F.

8. A device based on railway rail grounding detection test, characterized in that, include: The acquisition module is used to acquire a first railway track image, which is a railway track image taken by an aircraft. Image processing module, which is used to perform image processing on the first rail image to obtain the second rail image; An abnormal pixel extraction module is used to obtain a fourth rail image, which is a second rail image with abnormal pixel grayscale values. The judgment module is used to determine whether the abnormal pixel in the fourth rail image is in a first abnormal state or a second abnormal state based on the fifth rail image list. The first abnormal state is that there is a foreign object in the rail, and the second abnormal state is that the rail is rusted. The device is used to implement a method for detecting rail grounding as described in any one of claims 1-7.

9. A device for detecting railway grounding, characterized in that, The device includes: a processor and a memory storing computer program instructions; When the processor executes the computer program instructions, it implements the method for railway grounding detection test as described in any one of claims 1-7.

10. A computer storage medium, characterized in that, The computer storage medium stores computer program instructions, which, when executed by a processor, implement the method for rail grounding detection test as described in any one of claims 1-7.

Images