Railway tunnel drainage ditch state real-time monitoring method based on unmanned aerial vehicle

Abstract

The invention discloses a railway tunnel drainage ditch state real-time monitoring method based on an unmanned aerial vehicle. The method comprises the following steps: S1, realizing wireless data connection between the unmanned aerial vehicle inspection module and the mobile terminal through the image data transmission module; s2, a high-definition camera of the unmanned aerial vehicle inspection module collects image data of a railway tunnel drainage ditch; s3, transmitting the image data to the mobile terminal through the image data transmission module; s4, an image preprocessing module of the mobile terminal comprises a lightweight image processing framework, the image data are preprocessed, and then the preprocessed image data are sent to a model inference module of the mobile terminal; and S5, a model inference module comprises a pre-trained lightweight model, state inference is performed on the preprocessed image data, and a state inference result is displayed to a user through a human-computer interaction interface of the mobile terminal.

Description

Technical Field

[0001] The present invention belongs to the technical field of unmanned aerial vehicles, and in particular relates to a method for real-time monitoring of drainage ditch status in a railway tunnel based on an unmanned aerial vehicle. Background Art

[0002] Traditional drone inspection technology relies on back-end computers for image processing, model training and evaluation. This method requires powerful computing power and cannot be deployed on mobile terminals, which limits the application of drones in real-time monitoring and inspection integration, especially in railway tunnel drainage ditch status monitoring. How to improve the real-time and mobility of drone monitoring of railway tunnel drainage ditch status has become an urgent problem to be solved in this field. Summary of the invention

[0003] The present invention provides a method for real-time monitoring of the status of a drainage ditch in a railway tunnel based on an unmanned aerial vehicle, the method comprising the following steps:

[0004] S1. The drone inspection module is connected to the mobile terminal via the image data transmission module;

[0005] S2. The high-definition camera of the drone inspection module collects image data of the railway tunnel drainage ditch;

[0006] S3. The image data is transmitted to the mobile terminal via the image data transmission module;

[0007] S4. The image preprocessing module of the mobile terminal includes a lightweight image processing framework, which preprocesses the image data and then sends the preprocessed image data to the model inference module of the mobile terminal;

[0008] S5. The model inference module includes a pre-trained lightweight model, performs state inference on the preprocessed image data, and displays the state inference result to the user through the human-computer interaction interface of the mobile terminal.

[0009] Furthermore, the image data is grayscale image data of M×N pixels.

[0010] Further, preprocessing the image data specifically includes:

[0011] S41. Obtain the pixel value of the point (x, y) in the image data as I(x, y);

[0012] S42. Calculate the first intermediate processed pixel value I'(x, y) of the point (x, y),

[0013]

[0014] Among them, Ω is the neighborhood, s is the offset, h is the intensity coefficient, and exp represents the exponential operation;

[0015] S43. Calculate the second intermediate processed pixel value I" (x, y) of the point (x, y),

[0016] I”(x,y)=Σ s∈Ω I(x,y+s)G s (||s||)

[0017] Among them, G s () is the Gaussian kernel function, ||s|| is the distance from s to the center of the neighborhood;

[0018] S43. Calculate the pixel value O(x,y) of the midpoint (x,y) of the preprocessed output data,

[0019] O(x,y)=(exp(log(I”(x,y)))*H) γ

[0020] Where H is the high-pass filter, γ is the contrast adjustment gamma value, and log represents the logarithmic operation.

[0021] Further, the state inference of the preprocessed image data specifically includes:

[0022] S51. Calculate the first intermediate output O'(x,y),

[0023] O'(x,y)=∑ m ∑ n I(xm,yn)*K(m,n)

[0024] Among them, m is the offset of the convolution kernel in the x direction, n is the offset of the convolution kernel in the y direction, K is a lightweight convolution kernel, and its order is less than or equal to 3;

[0025] S52. Extract feature vector F,

[0026] F=[O'(1,1),O'(1,1),…,O'(1,N),O'(2,1),…,O'(M,N)]

[0027] S53. Calculate the second intermediate output z l ,

[0028] z l =W l ·F+b l

[0029] Among them, W l is the weight corresponding to the lth state type, which is b l The bias parameter corresponding to the lth state type;

[0030] S54. Calculate the probability pl ,

[0031]

[0032] Among them, k is the independent variable, L is the total number of state types;

[0033] S55. Perform state inference, and calculate the probability p corresponding to each of the L state types. l Sorting, the type corresponding to the highest ranking is used as the state inference result;

[0034] Furthermore, the state types are pre-trained, including normal, blocked, waterlogged, mineral crystallized, frozen, and damaged railway tunnel drainage ditches.

[0035] The present invention also relates to a computer program product, which includes a computer program, which is executed by a processor and is used to execute the real-time monitoring method for the status of a railway tunnel drainage ditch based on an unmanned aerial vehicle as described above.

[0036] The present invention also relates to a computer-readable storage medium, which is used to store a computer program. The computer program is executed by a processor to execute the real-time monitoring method for the status of a railway tunnel drainage ditch based on an unmanned aerial vehicle as described above.

[0037] The technical solution of the present invention fully considers the complexity of the model, the number of parameters, the computational cost and the inference speed, and can run efficiently on resource-constrained mobile devices, thereby improving the real-time and mobility of drone monitoring of the railway tunnel drainage ditch status. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 The present invention is a flow chart of a method for real-time monitoring of the status of a railway tunnel drainage ditch based on an unmanned aerial vehicle. DETAILED DESCRIPTION

[0039] The present invention will be further described below in conjunction with the accompanying drawings. The following examples are only used to more clearly illustrate the technical solution of the present invention, and cannot be used to limit the scope of protection of the present invention. It should be pointed out that the following detailed description is exemplary and is intended to provide further explanation of the present invention.

[0040] Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those generally understood by those of ordinary skill in the art to which the present invention belongs. It should be noted that the terms used herein are only for the purpose of describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0041] Embodiment 1 of the present invention relates to a method for real-time monitoring of the status of a railway tunnel drainage ditch based on an unmanned aerial vehicle, and the method comprises the following steps.

[0042] S1. The UAV inspection module is connected to the mobile terminal via the image data transmission module.

[0043] S2. The high-definition camera of the drone inspection module collects image data of the railway tunnel drainage ditch.

[0044] The image data is grayscale image data of M×N pixels.

[0045] S3. The image data is transmitted to the mobile terminal through the image data transmission module.

[0046] S4. The image preprocessing module of the mobile terminal includes a lightweight image processing framework, which preprocesses the image data and then sends the preprocessed image data to the model inference module of the mobile terminal.

[0047] The preprocessing of image data specifically includes:

[0048] S41. Obtain the pixel value of the point (x, y) in the image data as I(x, y).

[0049] S42. Calculate the first intermediate processed pixel value I'(x, y) of the point (x, y),

[0050]

[0051] Among them, Ω is the neighborhood, s is the offset, h is the strength coefficient, and exp represents the exponential operation.

[0052] S43. Calculate the second intermediate processed pixel value I" (x, y) of the point (x, y),

[0053] I”(x,y)=Σ s∈Ω I(x,y+s)G s (||s||)

[0054] Among them, Gs () is the Gaussian kernel function, ||s|| is the distance from s to the center of the neighborhood.

[0055] S43. Calculate the pixel value O(x,y) of the midpoint (x,y) of the preprocessed output data,

[0056] O(x,y)=(exp(log(I”(x,y)))*H) γ

[0057] Where H is the high-pass filter, γ is the contrast adjustment gamma value, and log represents the logarithmic operation.

[0058] S5. The model inference module includes a pre-trained lightweight model, performs state inference on the preprocessed image data, and displays the state inference result to the user through the human-computer interaction interface of the mobile terminal.

[0059] The state inference of the preprocessed image data specifically includes:

[0060] S51. Calculate the first intermediate output O'(x,y),

[0061] O'(x,y)=∑ m ∑ n I(xm,yn)*K(m,n)

[0062] Among them, m is the offset of the convolution kernel in the x direction, n is the offset of the convolution kernel in the y direction, and K is a lightweight convolution kernel whose order is less than or equal to 3.

[0063] S52. Extract feature vector F,

[0064] F=[O'(1,1),O'(1,1),…,O'(1,N),O'(2,1),…,O'(M,N)]

[0065] S53. Calculate the second intermediate output z l ,

[0066] z l =W l ·F+b l

[0067] Among them, W l is the weight corresponding to the lth state type, which is b l Bias parameter corresponding to the lth state type.

[0068] S54. Calculate the probability p l ,

[0069]

[0070] Among them, k is the independent variable and L is the total number of state types.

[0071] S55. Perform state inference, and calculate the probability p corresponding to each of the L state types. l Sorting, the type corresponding to the highest sorting is used as the state inference result.

[0072] The state types are pre-trained, and include, for example, normal, blocked, waterlogged, mineral crystallized, frozen, damaged, and the like types of railway tunnel drainage ditches.

[0073] Embodiment 2 of the present invention relates to a computer program product, which includes a computer program, and the computer program is executed by a processor to execute the real-time monitoring method of the railway tunnel drainage ditch status based on an unmanned aerial vehicle of Embodiment 1.

[0074] Embodiment 4 of the present invention relates to a computer-readable storage medium, which is used to store a computer program. The computer program is executed by a processor to execute the real-time monitoring method of the railway tunnel drainage ditch status based on an unmanned aerial vehicle of Embodiment 1.

[0075] In the process of traditional book selection and planning, planners are limited by their personal abilities and it is difficult to process and analyze large-scale data sets. The technical solution of the present invention uses advanced data preprocessing and analysis technologies to efficiently process massive data from multiple channels such as the Internet, libraries, and publishing houses, and provide more comprehensive and in-depth market insights. The existing technical methods often rely on the experience and intuition of planners when planning, which may lead to subjectivity and inconsistency in planning results. The technical solution of the present invention adopts deep learning and natural language processing technology based on ChatGPT, which can intelligently analyze market trends and reader needs and generate more objective and scientific planning schemes. It is difficult for traditional planning methods to provide customized services for different planning editors. The technical solution of the present invention can carry out personalized topic planning according to the specific needs and preferences of the planning editor to meet the unique needs of different users. Existing planning methods often find it difficult to break through the limitations of traditional thinking and lack innovation. The technical solution of the present invention can provide innovative topic planning schemes by integrating the latest artificial intelligence technology, and promote the innovative development of the publishing industry. In the existing planning methods, the feedback and iteration process of planning results are relatively slow and inefficient. The technical solution of the present invention supports rapid display and output of planning results, which facilitates planning editors to obtain feedback in a timely manner and perform iterative optimization. It is difficult to achieve continuous learning and optimization with traditional planning methods. The technical solution of the present invention has the ability of self-learning and optimization, and can continuously adjust and improve the planning algorithm according to the planning results and market feedback to ensure the continuous optimization of the planning scheme.

[0076] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the protection scope of the present invention.

Claims

1. A real-time monitoring method for railway tunnel drainage ditch status based on drones, characterized in that: The method comprises the following steps: S1. The drone inspection module is connected to the mobile terminal via the image data transmission module; S2. The high-definition camera of the drone inspection module collects image data of the railway tunnel drainage ditch; S3. The image data is transmitted to the mobile terminal via the image data transmission module; S4. The image preprocessing module of the mobile terminal includes a lightweight image processing framework, which preprocesses the image data and then sends the preprocessed image data to the model inference module of the mobile terminal; S5. The model inference module includes a pre-trained lightweight model, performs state inference on the preprocessed image data, and displays the state inference result to the user through the human-computer interaction interface of the mobile terminal.

2. The method for real-time monitoring of drainage ditch status in railway tunnels based on drones according to claim 1 is characterized in that: The image data is grayscale image data of M×N pixels.

3. The method for real-time monitoring of drainage ditch status in railway tunnels based on drones according to claim 2 is characterized in that: The preprocessing of image data specifically includes: S41. Obtain the pixel value of the point (x, y) in the image data as I(x, y); S42. Calculate the first intermediate processed pixel value I'(x, y) of the point (x, y), Among them, Ω is the neighborhood, s is the offset, h is the intensity coefficient, and exp represents the exponential operation; S43. Calculate the second intermediate processed pixel value I" (x, y) of the point (x, y), I”(x,y)=Σ s∈Ω I(x,y+s)G s (||s||) Among them, G s () is the Gaussian kernel function, ||s|| is the distance from s to the center of the neighborhood; S43. Calculate the pixel value O(x,y) of the midpoint (x,y) of the preprocessed output data, O(x,y)=(exp(log(I″(x,y)))*H) γ Where H is the high-pass filter, γ is the contrast adjustment gamma value, and log represents the logarithmic operation.

4. The method for real-time monitoring of the status of drainage ditches in railway tunnels based on drones according to claim 2 is characterized in that: The state inference of the preprocessed image data specifically includes: S51. Calculate the first intermediate output O′(x, y), O′(x,y)=∑ m ∑ n I(x-m,y-n)*K(m,n) Among them, m is the offset of the convolution kernel in the x direction, n is the offset of the convolution kernel in the y direction, K is a lightweight convolution kernel, and its order is less than or equal to 3; S52. Extract feature vector F, F=[O′(1,1),O′(1,1),...,O′(1,N),O′(2,1),...,O′(M,N)] S53. Calculate the second intermediate output z l , z l =W i ·F+b l Among them, W l is the weight corresponding to the lth state type, which is b l The bias parameter corresponding to the lth state type; S54. Calculate the probability p l , Among them, k is the independent variable, L is the total number of state types; S55. Perform state inference, and calculate the probability p corresponding to each of the L state types. l Sorting, the type corresponding to the highest sorting is used as the state inference result.

5. The method for real-time monitoring of the status of drainage ditches in railway tunnels based on drones according to any one of claims 1 to 4, characterized in that: The state types are pre-trained, including normal, blocked, waterlogged, mineral crystallized, frozen, and damaged railway tunnel drainage ditches.

6. A computer program product, characterized in that The computer program product includes a computer program, which is executed by a processor and is used to execute the real-time monitoring method for the status of a railway tunnel drainage ditch based on an unmanned aerial vehicle as described in any one of claims 1 to 5.

7. A computer-readable storage medium, characterized in that: The computer-readable storage medium is used to store a computer program, and the computer program is executed by a processor to execute the real-time monitoring method for the status of a railway tunnel drainage ditch based on an unmanned aerial vehicle as described in any one of claims 1 to 5.

Images