Method for detecting cable in working well

Abstract

The invention provides a method for detecting a cable in a working well. The method comprises the following steps: acquiring a to-be-detected image of a cable hole position layout in the working well; preprocessing the to-be-detected image to obtain a processed image; and inputting the processed image into a trained image detection model, and outputting a result reflecting the hole position layout difference degree between the processed image and the initial cable hole position layout image in the working well. Acquiring a scene image of the cable hole position layout in the working well as a to-be-detected image by using an underground detector front-end recognition device; meanwhile, real-time analysis is carried out at the front end through a trained image detection model; according to the method for detecting the cable in the working well, the result reflecting the hole position layout difference degree between the processed image and the initial hole position layout image of the cable in the working well can be output, the result includes but is not limited to hole position change, surrounding environment change and the like, the detection accuracy of cable identification can be improved, and the inspection efficiency of a cable line is remarkably improved.

Description

Technical Field

[0001] This application relates to the technical field of cable detection in manholes, and specifically relates to a method for detecting cables in manholes. Background Art

[0002] With the rapid development of all aspects of society in our country, the demand for electricity consumption and supporting power facilities has also increased sharply. Due to the advantages of small floor area, strong anti-interference ability, and high safety of the underground cable transmission power supply method, the traditional overhead power supply lines have been gradually replaced by this new power supply method. However, at the same time, new maintenance safety problems have emerged. The cable manhole is the intersection of cable interfaces. As an intermediate connection link for power transmission, it is an important channel for maintenance and operation.

[0003] In recent years, the underground maintenance safety problems of cable manholes have been particularly prominent. Safety accidents are more likely to occur during maintenance operations, posing serious safety hazards to the lives of operators and the maintenance of the power grid, which has attracted the attention and concern of the power maintenance department. Summary of the Invention

[0004] This application provides a method for detecting cables in manholes, which can improve the detection accuracy of cable identification and significantly improve the inspection efficiency of cable lines. The specific solution is as follows:

[0005] This application provides a method for detecting cables in manholes, and the method includes:

[0006] Obtain a to-be-detected image of the cable hole layout in the manhole;

[0007] Preprocess the to-be-detected image to obtain a processed image;

[0008] Input the processed image into a trained image detection model, and output a result reflecting the hole layout difference degree between the processed image and the initial cable hole layout image in the manhole. The result of the hole layout difference degree is used to detect the cable hole layout in the manhole.

[0009] Optionally, obtaining a to-be-detected image of the cable hole layout in the manhole includes:

[0010] Use an image capture device to capture pictures of the cable hole layout in the manhole from multiple angles and directions as the to-be-detected image.

[0011] Optionally, the training method of the trained image detection model includes:

[0012] Build a neural network model;

[0013] Obtain an initial image of the cable hole layout in the manhole when cables are laid in the manhole;

[0014] Perform grayscale processing on the initial image, and use the grayscale-processed initial image as the reference data set;

[0015] Obtain a training image of the cable hole layout in the manhole after laying cables in the manhole;

[0016] Perform grayscale processing on the training image, and use the grayscale-processed training image as the input data set;

[0017] Construct a validation data set based on the difference between the hole layout in the initial image and the hole layout in the training image;

[0018] Use the input data set, the reference data set, and the validation data set to train the neural network model to obtain a trained image detection model.

[0019] Optionally, constructing a validation data set based on the difference between the hole layout in the initial image and the hole layout in the training image includes:

[0020] Obtain the grayscale image of the initial image and the grayscale image of the training image;

[0021] Subtract the grayscale values of the grayscale image of the initial image and the grayscale image of the training image, and splice them to obtain the spliced grayscale image, and use the spliced grayscale image as the validation data set.

[0022] Optionally, the method further includes: using multiple groups of the input data set and the validation data set to train and calibrate the trained image detection model.

[0023] Optionally, preprocessing the image to be detected to obtain a processed image includes:

[0024] Determine the position of the target feature in the image to be detected; and determine the position of the target feature in the initial image; wherein, the target feature is: cable holes and / or cables in the manhole;

[0025] Based on the position of the target feature in the image to be detected and the position of the target feature in the initial image, screen and obtain the image to be detected that meets the preset conditions to obtain the screened image to be detected;

[0026] Perform exposure and grayscale processing on the screened image to be detected to obtain an image to be detected with a grayscale value matching that of the initial image, and crop the image to be detected with a matching grayscale value based on the initial image to obtain a processed image.

[0027] Optionally, the grayscale processing includes:

[0028] Transform the initial image into a grayscale image and obtain the image mean of the grayscale image;

[0029] Use the initial image with the image mean within a preset range as the image to be adjusted, adjust the brightness and darkness of the image to be detected, and obtain the image to be detected with a matched grayscale value.

[0030] Optionally, input the processed image into a trained image detection model, and output the result reflecting the difference in hole layout between the processed image and the initial cable hole layout image in the manhole, including:

[0031] Input the processed image into a trained image detection model;

[0032] The trained image detection model uses the opencv edge detection algorithm to detect the position of the target feature in the processed image, and uses the edge algorithm to perform edge filling on the contour of the position of the target feature to obtain the image to be compared;

[0033] Compare the contour of the target feature of the image to be compared with the contour of the target feature of the initial image, mark the difference between the two, and output the result reflecting the difference in hole layout between the processed image and the initial cable hole layout image in the manhole.

[0034] Optionally, using the edge algorithm to perform edge filling on the contour of the position of the target feature includes:

[0035] Randomly divide the processed image to obtain multiple image fragments;

[0036] Perform edge filling on the contour of the position of the target feature in each of the image fragments;

[0037] Combine the edge-filled image fragments to obtain the image to be compared.

[0038] Optionally, output the result reflecting the difference in hole layout between the processed image and the initial cable hole layout image in the manhole, where the result of the layout difference includes at least one of: hole occupancy status, hole displacement status, diameter of the cable, cable cut-off status.

[0039] Compared with the prior art, the present application has the following advantages:

[0040] A method for detecting cables in a manhole provided by this application includes: obtaining a to-be-detected image of the cable hole layout in the manhole; preprocessing the to-be-detected image to obtain a processed image; inputting the processed image into a trained image detection model, and outputting a result reflecting the hole layout difference degree between the processed image and the initial cable hole layout image in the manhole. Using the front-end recognition device of an underground detector to obtain a scene image of the cable hole layout in the manhole as the to-be-detected image; at the same time, performing real-time analysis through the trained image detection model at the front end; and a result reflecting the hole layout difference degree between the processed image and the initial cable hole layout image in the manhole can be output, including but not limited to hole position changes, surrounding environment changes, etc. The method for detecting cables in a manhole provided by this application can improve the detection accuracy of cable identification and significantly enhance the inspection efficiency of cable lines by detecting the downhole image through the front-end recognition device of the underground detector and performing real-time analysis on the downhole image. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 is a schematic flowchart of the method for detecting cables in a manhole provided by the first embodiment of this application;

[0042] Figure 2 is a schematic flowchart of the training process of the image recognition model provided by the embodiment of this application;

[0043] Figure 3 is an example diagram of a device for detecting cables in a manhole provided by the embodiment of this application;

[0044] Figure 4 is a schematic diagram of an electronic device provided by the third embodiment of this application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0045] In order to enable those skilled in the art to better understand the technical solutions of this application, the following describes this application clearly and completely in conjunction with the accompanying drawings in the embodiments of this application. However, this application can be implemented in many other ways different from the following description. Therefore, based on the embodiments provided by this application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of this application.

[0046] It should be noted that the terms "first", "source domain", "third", etc. in the claims, the description and the drawings of the present application are used to distinguish similar objects, and are not used to describe a specific order or sequence. Data used in this way can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than that shown or described herein. In addition, the terms "comprising", "having" and their variants are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device comprising a series of steps or units need not be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0047] To facilitate the understanding of the embodiments of the present application, the application background of the embodiments is described.

[0048] With the acceleration of the urbanization process, the construction and maintenance of underground infrastructure have become increasingly important. As an important part of power transmission and communication, the safety and reliability of underground cables directly affect the normal operation of the city. Therefore, it is often necessary to repair underground cables. To facilitate construction workers to quickly find the buried cables, cable burial marks are usually set on the ground. However, over time, the cable burial marks set on the ground may be damaged by external forces, making it impossible for maintenance personnel to find them, resulting in construction workers being unable to accurately find the buried position of the cables, which brings great difficulties to the maintenance of underground cables.

[0049] The related technology usually involves construction workers selecting a general location for excavation according to previous design drawings, and the excavation area is relatively large to increase the probability of finding the cables. This method is not only inefficient but also wastes manpower.

[0050] To solve the above problems, the embodiments of the present application provide a method, a device, an electronic device and a computer-readable storage medium for detecting cables in a manhole. It can accurately find the buried position of the cables when the cable burial marks set on the ground cannot be found, so as to facilitate the maintenance of underground cables.

[0051] The first embodiment of the present application provides a method for detecting cables in a manhole. The execution subject of this method can be a controller, which can be a device with control functions and data transmission functions such as a processor, a microprocessor, a desktop computer, a laptop computer, a smart mobile terminal, a server, etc. The controller can be installed in a radio frequency detection device or can be a device installed outside the radio frequency detection device and communicatively connected to the radio frequency detection device.

[0052] As Figure 1 shown, a method for detecting cables in a manhole provided by the first embodiment of the present application includes the following steps S110 to step S130.

[0053] S110. Obtain the image to be detected of the cable hole layout in the manhole.

[0054] In this embodiment, a device with a camera or a separate camera can be used to capture the image to be detected of the cable hole layout in the manhole. Devices with cameras include, for example, cameras, cameras, and smartphones with cameras. This application does not make any limitations.

[0055] Further, when capturing the above-mentioned image to be detected, first open the manhole, then place the camera into the manhole, and capture the cable hole layout in the manhole to obtain the aforementioned image to be detected.

[0056] In this embodiment, the camera or the device with a camera can be used to capture the cable hole layout in the manhole from multiple angles to obtain images to be detected that can display the cable hole layout from different angles and directions.

[0057] Further, the image to be detected needs to include information such as hole position changes, surrounding environment, and the current layout of the cables. Displaying the above information in the image to be detected can be used to identify whether the hole is occupied, the thickness of the cable, and whether the cable is cut during the subsequent image recognition process.

[0058] S120. Preprocess the image to be detected to obtain the processed image.

[0059] In this embodiment, in order to more accurately identify and compare the image to be detected, after obtaining the image to be detected, it needs to be preprocessed. The processing methods include but are not limited to exposure, grayscale processing, cropping, and whether the photographing position is required to be the same. Grayscale processing usually refers to the process of converting a color image into a grayscale image. A grayscale image is an image that only contains grayscale levels (different brightness levels between black and white) and does not contain any color information. Grayscale images are easier to process than color images in many image processing tasks because they remove color information and reduce computational complexity. The image to be detected also needs to be cropped so that its size matches that of the initial image, and the hole positions on the image to be detected match the hole positions on the initial image. In this way, it is more convenient for subsequent image comparison.

[0060] Further, in the preprocessing of the image to be detected, an image processing software can be used to pre-match the hole positions between the initial image and the obtained image to be detected, and then crop the image so that the external shape of the image to be detected matches that of the initial image. In this way, it is convenient for the subsequent comparison between the image to be detected and the initial image.

[0061] S130. Input the processed image into a trained image detection model, and output a result reflecting the hole layout difference between the processed image and the initial cable hole layout image in the working shaft, wherein the hole layout difference result is used to detect the cable hole layout in the working shaft.

[0062] In this embodiment, the trained image detection model can be used to identify the processed image, and the before and after differences between the two underground pictures can be output, including changes in hole positions and changes in the surrounding environment. The method of detecting underground images by the underground detector front-end recognition device provided in this application can improve the detection accuracy of cable identification and significantly improve the inspection efficiency of cable lines.

[0063] Furthermore, in step S130, the training method of the trained image detection model includes: steps S1301 to S1307:

[0064] S1301. Build a neural network model.

[0065] In this embodiment, the neural network model may use the original YOLOv5 network model, or other neural network models that can be used for image comparison analysis, which is not limited in this application.

[0066] S1302. Acquire an initial image of the layout of cable holes in the work pit when the cables are laid in the work pit;

[0067] In this embodiment, the initial image is designed when the cable is first constructed, and the initial image can be obtained from the construction record. The initial image shows the cable hole layout after the initial construction is completed. Of course, the above initial image can also be a specific time point, for example, an image one year after construction, etc. In this way, the subsequent inspection work will be compared with the time node of the determined initial image. The user can set it according to the actual situation, and this application does not make specific restrictions.

[0068] For example, during construction, three cable holes are designed, namely hole 1, hole 2, and hole 3; cable 1 is set in hole 1; cable 2 is set in hole 2; and cable 3 is set in hole 3; that is, the above information is displayed in the initial image.

[0069] S1303. Perform grayscale processing on the initial image, and use the grayscale processed initial image as a benchmark data set.

[0070] In this embodiment, as a benchmark data set, it means that all newly acquired images to be detected are compared with the above benchmark data set as a benchmark, and all differences are compared with the aforementioned benchmark data. The above benchmark data can be set by the user according to actual usage requirements.

[0071] For example, after grayscale processing the initial image, data such as the aperture diameters of hole position 1, hole position 2, and hole position 3, as well as the cable diameters of cable 1, cable 2, and cable 3 can be clearly obtained, serving as a benchmark for subsequent comparison with the image to be detected.

[0072] S1304. Obtain a training image of the cable hole position layout in the manhole after threading cables in the manhole.

[0073] In this embodiment, to facilitate the training of the model, an image of the cable hole position layout in the manhole within a time period after the time node of the initial image needs to be obtained.

[0074] S1305. Perform grayscale processing on the training image and use the grayscale-processed training image as the input data set.

[0075] In this embodiment, the obtained training image needs to be grayscale processed to obtain an image that can reflect the cable hole position layout in the manhole at a certain time point or time period after the time node of the initial image. This image is used as the input data set for model training, and the model is used to compare it with the reference data set to achieve the function of training the model parameters.

[0076] S1306. Construct a validation data set based on the difference between the hole position layout in the initial image and the hole position layout in the training image.

[0077] In this embodiment, the validation data set needs to be obtained by the user himself, that is, by subtracting the grayscale-processed reference data set from the grayscale-processed input data set, the above-mentioned validation data set can be obtained.

[0078] S1307. Use the input data set, the reference data set, and the validation data set to train the neural network model to obtain a trained image detection model.

[0079] In this embodiment, the image detection model can be trained through the input data set, the output data set, and the validation data set. Of course, after the initial training of the image detection model, preliminary model parameters can be obtained, and the image training model also needs to be calibrated to improve the accuracy of the image training model.

[0080] In a specific implementation manner, obtaining the image to be detected of the cable hole position layout in the manhole includes: using an image capture device to capture pictures of the cable hole position layout in the manhole from multiple angles and directions as the image to be detected.

[0081] In a specific embodiment, constructing a validation dataset based on the difference between the hole layout in the initial image and the hole layout in the training image includes: obtaining the grayscale images of the initial image and the training image; subtracting the grayscale values of the grayscale image of the initial image and the grayscale image of the training image, and splicing them to obtain the spliced grayscale image, and using the spliced grayscale image as the validation dataset. Changing the grayscale of the two images, then graying the subtracted image, and splicing three grayscale images for model training.

[0082] In a specific embodiment, the method further includes: calibrating the trained image detection model by using multiple groups of the input dataset and the validation dataset.

[0083] In a specific embodiment, preprocessing the image to be detected to obtain a processed image includes:

[0084] determining the position of the target feature in the image to be detected; and determining the position of the target feature in the initial image; wherein, the target feature is: the cable holes and / or cables in the manhole.

[0085] Based on the position of the target feature in the image to be detected and the position of the target feature in the initial image, screening and obtaining the images to be detected that meet the preset conditions to obtain the screened images to be detected;

[0086] Performing exposure and grayscale processing on the screened images to be detected to obtain the images to be detected with grayscale values matching those of the initial image, and cropping the images to be detected with matching grayscale values based on the initial image to obtain the processed image.

[0087] In a specific embodiment, the grayscale processing includes:

[0088] transforming the initial image into a grayscale image and obtaining the image mean of the grayscale image;

[0089] Taking the initial image with the image mean within the preset range as the image to be adjusted, adjusting the brightness and darkness of the image to be detected, and obtaining the image to be detected with matching grayscale values.

[0090] Specifically: Adjust the grayscale of the image to be detected to obtain an image to be detected with a matching grayscale value. Specifically, obtain the maximum illumination distance and the division distance length in the image to be detected. The maximum illumination distance is d1 = 255 - M, and the division distance length is s, where s is a preset value that can be determined through experiments. According to the maximum illumination distance and the division distance length, obtain the adjustment quantity, and round down the adjustment quantity n = [d1 / s]. According to the division distance length and the adjustment quantity, adjust the pixels of the image to be detected to obtain an image to be detected with a matching grayscale value, that is, perform n image pixel adjustments on the image to be adjusted. Each time the adjusted pixel value pixel = pixel + s, and after n adjustments, an image to be detected with a matching grayscale value is obtained.

[0091] Further, it is necessary to use the grayscale value of the initial image as an adjustment reference to adjust the grayscale value of the image to be detected.

[0092] In a specific implementation manner, input the processed image into a trained image detection model, and output a result reflecting the hole layout difference degree between the processed image and the initial cable hole layout image in the manhole, including:

[0093] Input the processed image into a trained image detection model;

[0094] The trained image detection model uses the opencv edge detection algorithm to detect the position of the target feature in the processed image, and uses the edge algorithm to perform edge filling on the contour of the position of the target feature to obtain an image to be compared;

[0095] In this embodiment, the opencv edge detection algorithm can use the Canny edge detection algorithm, noise removal (using Gaussian blur); calculate the gradient intensity and direction of the image; non-maximum suppression (remove unnecessary points on the edge); double threshold processing to detect potential edges; in OpenCV, use the cv2.Canny() function to implement the Canny edge detection.

[0096] Compare the contour of the target feature of the image to be compared with the contour of the target feature of the initial image, mark the differences between the two, and output a result reflecting the hole layout difference degree between the processed image and the initial cable hole layout image in the manhole.

[0097] In a specific implementation manner, using the edge algorithm to perform edge filling on the contour of the position of the target feature includes:

[0098] Randomly divide the processed image to obtain multiple image fragments;

[0099] Perform edge filling on the contour of the position of the target feature in each of the image fragments;

[0100] Combine the image fragments after edge filling to obtain the image to be compared.

[0101] In a specific embodiment, output a result reflecting the hole position layout difference degree between the processed image and the initial cable hole position layout image in the manhole, where the result of the layout difference degree includes at least one of: the hole position occupancy state, the hole position displacement state, the diameter of the cable, and the cable cutting state.

[0102] The second embodiment of the present application also provides a cable position detection device corresponding to the manhole cable detection method embodiment provided in the above first embodiment. Since the device embodiment is basically similar to the method embodiment, the description is relatively simple. For the details of the relevant technical features and the achieved effects, please refer to the corresponding description of the manhole cable detection method embodiment provided above. As Figure 3 described, the manhole cable detection device 300 provided by the present application includes:

[0103] An acquisition unit 301, configured to acquire an image to be detected of the cable hole position layout in the manhole;

[0104] A preprocessing unit 302, configured to preprocess the image to be detected to obtain a processed image;

[0105] A comparison unit 303, configured to input the processed image into a trained image detection model, and output a result reflecting the hole position layout difference degree between the processed image and the initial cable hole position layout image in the manhole, and the result of the hole position layout difference degree is used to detect the cable hole position layout in the manhole.

[0106] The third embodiment of the present application also provides an electronic device embodiment corresponding to the manhole cable detection method embodiment provided in the above first embodiment. The following description of the electronic device embodiment is merely illustrative. The electronic device embodiment is as follows:

[0107] Please refer to Figure 4 to understand the above electronic device, Figure 4 which is a schematic diagram of the electronic device. The electronic device provided in this embodiment includes: a processor 401, a memory 402, a communication bus 403, and a communication interface 404;

[0108] The memory 402 is used to store computer instructions for data processing. When the computer instructions are read and executed by the processor 401, the following steps are performed:

[0109] Acquire an image to be detected of the cable hole position layout in the manhole;

[0110] Preprocess the image to be detected to obtain a processed image;

[0111] Input the processed image into a trained image detection model, and output a result reflecting the hole layout difference degree between the processed image and the initial cable hole layout image in the manhole. The result of the hole layout difference degree is used to detect the cable hole layout in the manhole.

[0112] The fourth embodiment of this application also provides a computer-readable storage medium for implementing the above cable detection method in a manhole. The embodiment of the computer-readable storage medium provided in this application is described relatively simply. For the corresponding description of the relevant part, please refer to the corresponding description of the above method embodiment. The following described embodiments are only illustrative.

[0113] The computer-readable storage medium provided in this embodiment stores computer instructions, and when the instructions are executed by a processor, the following steps are implemented:

[0114] Obtain an image to be detected of the cable hole layout in the manhole;

[0115] Preprocess the image to be detected to obtain a processed image;

[0116] Input the processed image into a trained image detection model, and output a result reflecting the hole layout difference degree between the processed image and the initial cable hole layout image in the manhole. The result of the hole layout difference degree is used to detect the cable hole layout in the manhole.

[0117] In a typical configuration, a computing device includes one or more processors (CPUs), an input / output interface, a network interface, and a memory.

[0118] The memory may include non-permanent memory in the computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of, for example, read-only memory (ROM) or flash memory (flash RAM). The memory is an example of a computer-readable medium.

[0119] 1. A computer-readable medium includes both permanent and non-permanent, removable and non-removable media that can store information by any method or technology. The information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transitory medium that can be used to store information accessible by a computing device. As defined herein, a computer-readable medium does not include transitory media such as modulated data signals and carrier waves.

[0120] 2. Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memory, CD-ROM, optical memory, etc.) that contain computer-usable program code.

[0121] Although the present application is disclosed above in preferred embodiments, it is not intended to limit the present application. Any person skilled in the art can make possible changes and modifications without departing from the spirit and scope of the present application. Therefore, the protection scope of the present application should be determined by the scope defined in the claims of the present application.

Claims

1. A method for detecting cables in a working well, characterized in that: The method comprises: Obtain the image to be inspected of the cable hole layout in the working well; Preprocessing the image to be detected to obtain a processed image; The processed image is input into a trained image detection model, and a result reflecting the difference in hole layout between the processed image and the initial cable hole layout image in the working shaft is output, and the result of the hole layout difference is used to detect the cable hole layout in the working shaft.

2. The cable detection method in a working well according to claim 1, characterized in that: Obtain the image to be inspected of the cable hole layout in the manhole, including: An image capturing device is used to capture pictures of the cable hole layout in the working well from multiple angles and directions as the image to be detected.

3. The cable detection method in a working well according to claim 1, characterized in that: The training method of the trained image detection model includes: Build a neural network model; Acquire an initial image of the layout of cable holes in the work pit when laying cables in the work pit; Performing grayscale processing on the initial image, and using the grayscale processed initial image as a benchmark data set; Acquire a training image of the layout of cable holes in the work well after the cables are inserted into the work well; Performing grayscale processing on the training image, and using the grayscale processed training image as an input data set; Building a validation data set based on the difference between the hole layout in the initial image and the hole layout in the training image; The neural network model is trained using the input data set, the benchmark data set, and the verification data set to obtain a trained image detection model.

4. The cable detection method in a working well according to claim 3, characterized in that: Constructing a verification data set based on the difference between the hole layout in the initial image and the hole layout in the training image comprises: Obtain a grayscale image of the initial image and a grayscale image of the training image; The grayscale values ​​of the grayscale image of the initial image and the grayscale values ​​of the grayscale image of the training image are subtracted and spliced ​​to obtain the spliced ​​grayscale image, and the spliced ​​grayscale image is used as the verification data set.

5. The cable detection method in a working well according to claim 3, characterized in that: The method also includes: calibrating the trained image detection model using multiple sets of the input data sets and the verification data sets for training.

6. The cable detection method in a working well according to claim 3, characterized in that: Preprocessing the image to be detected to obtain a processed image includes: Determine the position of the target feature in the image to be detected; and determine the position of the target feature in the initial image; wherein the target feature is: a cable hole and / or a cable in a working well; Based on the position of the target feature in the image to be detected and the position of the target feature in the initial image, screening and acquiring the image to be detected that meets the preset conditions, to obtain the screened image to be detected; The screened image to be detected is exposed and gray-scale processed to obtain an image to be detected whose grayscale value matches the grayscale value of the initial image, and the image to be detected whose grayscale value matches is cropped based on the initial image to obtain a processed image.

7. The method for detecting cables in a working well according to claim 6, characterized in that: The grayscale processing includes: Convert the initial image into a grayscale image and obtain an image mean of the grayscale image; An initial image whose image mean is within a preset range is used as the image to be adjusted, and the brightness of the image to be detected is adjusted to obtain an image to be detected with matching grayscale values.

8. The method for detecting cables in a working well according to claim 6, characterized in that: The processed image is input into a trained image detection model, and a result reflecting the difference between the processed image and the initial cable hole layout image in the working well is output, including: Inputting the processed image into a trained image detection model; The trained image detection model uses the opencv edge detection algorithm to detect the position of the target feature in the processed image, and uses the edge algorithm to fill the edge of the contour of the position of the target feature to obtain the image to be compared; The contour of the target feature of the image to be compared is compared with the contour of the target feature of the initial image, and the difference between the two is marked, and the result reflecting the difference in hole layout between the processed image and the initial cable hole layout image in the working well is output.

9. The method for detecting cables in a working well according to claim 8, characterized in that: Filling the edge of the contour of the position of the target feature by using an edge algorithm includes: Randomly dividing the processed image to obtain a plurality of image fragments; Filling the edges of the contours of the positions of the target features in each of the image fragments; The image fragments after each edge filling are combined to obtain the image to be compared.

10. The method for detecting cables in a working well according to claim 8, characterized in that: The output reflects the result of the hole layout difference between the processed image and the initial cable hole layout image in the working well, wherein the layout difference result includes at least one of: hole occupancy status, hole displacement status, cable diameter, and cable cut status.

Images