Anti-seepage monitoring method for cofferdam in tidal zone bearing platform construction

Abstract

The invention provides a cofferdam anti-seepage monitoring method for tidal zone bearing platform construction. According to the method, real-time monitoring and early warning of the cofferdam state are achieved. Firstly, an acquisition module acquires high-definition image data of the surface of the cofferdam, cofferdam water pressure and tide data, and time stamps are ensured to be synchronous; then, using computer vision and a deep learning algorithm to identify crack, attitude deviation and deformation conditions in the image; predicting future water pressure change by using a deep learning model in combination with the tide data and the water pressure data; and when the detection result exceeds a preset threshold value, an early warning mechanism is triggered immediately. In addition, all data are stored in a database, and model parameters are updated regularly to improve accuracy. According to the method, multi-source data are fused, possible factors of cofferdam penetration are detected by adopting a computer vision technology and a deep learning technology, and the cofferdam anti-seepage monitoring accuracy and the dynamic prediction capability are remarkably improved.

Description

Technical Field

[0001] The invention relates to the fields of water conservancy engineering and marine engineering, and specifically belongs to the technical field of cofferdam construction monitoring and anti-seepage management. Background Art

[0002] Tidal zone cofferdams are common temporary structures in water conservancy projects and marine engineering projects. Their main function is to provide a dry working environment for construction in the tidal impact area. Due to the special geographical conditions of the tidal zone, the cofferdam not only needs to withstand the external water pressure, but also needs to cope with the periodic changes in tidal height, wave height and current speed, which poses great challenges to the stability and impermeability of the cofferdam.

[0003] In practical applications, the anti-seepage performance of the cofferdam is crucial to the smooth progress of the project and the safety of construction. If the cofferdam leaks, it may cause water to invade the internal construction area, or even cause the cofferdam to collapse, seriously threatening the safety of construction workers and increasing project costs. Therefore, how to conduct real-time and accurate anti-seepage monitoring and status assessment of the cofferdam is an important technical problem that needs to be solved in the current engineering field.

[0004] Traditional cofferdam monitoring methods rely on manual inspection or simple single sensor monitoring. Manual inspection is not only time-consuming and laborious, but also has subjective judgment errors, and a single sensor cannot fully reflect the overall status of the cofferdam. With the rapid development of computer vision and deep learning technology, new solutions have been provided for cofferdam anti-seepage monitoring. Summary of the invention

[0005] The present invention proposes a method for monitoring the anti-seepage of cofferdams in tidal area foundation construction. The method can not only monitor abnormal conditions on the surface of the cofferdam through image recognition technology, but also detect pressure changes outside the cofferdam using water pressure sensors and predict future water pressure changes through artificial intelligence algorithms, thereby more comprehensively evaluating the anti-seepage status of the cofferdam.

[0006] To this end, the technical solution adopted in the present invention is as follows: S1. Collect high-definition image data of the cofferdam surface, collect the time series of the cofferdam water pressure, and collect tidal data, including tidal height, wave height and current speed. The timestamps of tidal data and cofferdam water pressure data are synchronized; S2. Use computer vision and deep learning algorithms to identify the collected high-definition image data and determine whether there are factors that cause leakage in the cofferdam, including crack detection, posture deviation detection, and deformation detection; Detect the length and width of the cofferdam cracks, the rotation angle and translation of the cofferdam posture deviation, and whether the cofferdam is deformed; S3, build a deep learning model, use the collected tidal data and cofferdam water pressure data to predict the cofferdam water pressure in the future; S4. Setting dangerous thresholds for crack length and width, rotation angle and translation, and immediately triggering the warning mechanism when the crack detection and posture deviation detection results exceed the set thresholds or the deformation detection results are convex and concave; Set the water pressure danger threshold of the cofferdam to immediately trigger the early warning mechanism; S5. Establish a cofferdam status database, and store the high-definition image data of the cofferdam and the cofferdam status detection results in the database. A cofferdam water pressure database is established to store water pressure sensor data, local tidal data, and future cofferdam water pressure prediction results in the database, and the deep learning model parameters in S3 are regularly updated with the latest water pressure data and tidal data.

[0007] Furthermore, before the high-definition image data is recognized, the image data is preprocessed by grayscale, smoothing filtering and histogram equalization. The grayscale is specifically as follows: Grayscale is to average the three color channel (RGB) values ​​in the original image to get a grayscale value. The formula is:

[0008] in, Represents the grayscale value of the pixel in the i-th row and j-th column of the grayscale image. represents the R value of the i-th row and j-th column of the original high-definition image, represents the G value of the i-th row and j-th column of the original high-definition image, represents the B value of the i-th row and j-th column of the original high-definition image; Furthermore, the smoothing filter uses a Gaussian function as a weight to calculate the weighted average of a pixel and its neighboring pixels. The shape of the Gaussian function is controlled by its standard deviation σ. For a given pixel position (i, j), the gray value SF_Gray(i, j) after smoothing filtering is calculated by the following formula:

[0009] Among them, W is the neighborhood window centered at (i, j), and the window size is k*l , is the value of the Gaussian function, defined as:

[0010] Furthermore, the histogram equalization is to enhance the contrast of the image by adjusting the histogram distribution of the image, and its mapping function uses the cumulative distribution function, and the formula is:

[0011] Where L is the total number of gray levels, M and N are the length and width of the image respectively, is the number of pixels at the (i,j)th gray level after histogram equalization.

[0012] Furthermore, the specific steps of crack detection are: 1) For the preprocessed image Each pixel in , respectively with G i and G j Perform convolution to obtain the horizontal gradient approximation I i and the vertical gradient approximation I j , the formula is as follows,

[0013] in represents a two-dimensional convolution operation, G i and G j is the Sobel operator; 2) Calculate the gradient magnitude and gradient direction of each pixel, the gradient magnitude represents the strength of the edge, while the gradient direction Represents the direction of the edge, the formula is as follows,

[0014] 3) Perform non-maximum suppression on the gradient magnitude image M(i, j), specifically Quantize to one of the four main directions, namely horizontal, vertical, diagonal, and anti-diagonal. According to the quantized direction, the closest neighboring pixel in the gradient direction can be found. If the gradient magnitude of the pixel (i, j) is greater than the gradient magnitude of the neighboring pixel, the gradient magnitude of the pixel is retained, otherwise it is set to 0; 4) Perform threshold processing on the gradient magnitude image M(i,j), select a fixed threshold T, and regard all pixels with gradient magnitude greater than T as edge pixels, and other pixels as non-edge pixels; 5) Use the expansion operation to connect the fractured crack edges to form a complete crack outline. Specifically, select a rectangular structural element and align the origin of the structural element with For each pixel in the image, determine whether the area covered by the structural element intersects with the edge pixels in the image. If so, set the pixel value of the pixel to the average value of all covered edge pixels; if not, keep the original value of the pixel. The expanded image is represented as IF_Gray; 6) Select a fixed threshold T b , convert IF_Gray into a binary image, that is, greater than T b The pixel of is set to 255, otherwise it is set to 0; Use the skeleton extraction algorithm to reduce the lines of the binary image to a single pixel width, calculate the pixel length of the crack skeleton, and then convert the pixel length into the actual physical size according to the image resolution and the scale at the time of shooting to obtain the crack length; The pixel width of the crack in the image is calculated using the orthogonal skeleton method, and then the pixel distance is converted into the actual physical size according to the image resolution and the scale at the time of shooting to obtain the crack width.

[0015] Furthermore, the posture deviation detection is specifically to use the CNN model to identify the data after preprocessing Image, the model consists of 3 convolutional layers, 1 pooling layer, 1 fully connected layer and 1 output layer in sequence. The number of neurons in the output layer is 6, representing the rotation angles of the cofferdam around the x, y and z axes in three-dimensional space and the translation in the x, y and z directions respectively.

[0016] Furthermore, the deformation detection is specifically to use the CNN model to identify the data after preprocessing Image, the model consists of 3 convolutional layers, 1 pooling layer, 1 fully connected layer and 1 classification layer. The number of neurons in the classification layer is 3, representing normal, convex and concave cofferdams respectively.

[0017] Furthermore, the main body of the deep learning model of step S3 is a Bi-GRU structure, and the Bi-GRU includes a forward GRU structure and a backward GRU structure. The forward GRU structure can be expressed as: Forward update gate:

[0018] in, is the sigmoid activation function, is the weight matrix of the update gate, is the hidden state at the previous time step, is the tide data of the current time step, is the bias term, Forward reset gate:

[0019] in, is the weight matrix of the reset gate, is the bias term, Forward to new candidate state:

[0020] in, is the weight matrix of the new candidate state, is the bias term, Represents element-wise multiplication.

[0021] Update the hidden state forward:

[0022] in, is the hidden state at the current time step, The backward GRU structure is exactly the same as the forward one, and the final Bi-GRU output is,

[0023] in, is the hidden state of the forward GRU, is the hidden state of the inverse GRU, The Bi-GRU structure is followed by a fully connected layer.

[0024] Furthermore, the input to the deep learning model is a fixed time window T s The time series of tide height, wave height, current speed and cofferdam water pressure are output as lag time A fixed time window T s The water pressure time series, To find the best time window, select different time window sizes T s Verify to find the best lag time , select different lag times for verification, The mean square error (MSE) of the deep learning model under different time windows and different lag times was compared to determine the optimal time window and lag time, and the time series of real-time tidal data and cofferdam water pressure data were input into the model to predict future Water pressure data of the cofferdam after time.

[0025] Compared with the prior art, the advantages of the present invention are: 1. The present invention improves the accuracy of anti-seepage monitoring of cofferdams. By combining high-definition images with water pressure sensor data, it not only monitors the changes in the external structure of the cofferdam, but also reflects the water pressure changes in the underwater part in real time, and comprehensively covers the key areas of anti-seepage monitoring of cofferdams. The present invention uses computer vision technology and deep learning technology to process and identify images, and can efficiently and accurately detect cracks, deformations and posture deviations of cofferdams, ensuring early detection of anti-seepage problems.

[0026] 2. The present invention introduces the Bi-GRU structural model and utilizes the forward and backward features of the time series to comprehensively capture the dynamic changes of tides and water pressure, significantly improve the accuracy of water pressure prediction, and regularly updates the deep learning model parameters based on the latest data to ensure that the prediction results always maintain high efficiency and accuracy.

[0027] 3. When the detection result exceeds the set threshold, the present invention can automatically trigger an early warning and notify relevant personnel to take repair or reinforcement measures in time, reduce manual intervention, and save maintenance costs. The present invention builds a comprehensive cofferdam status and water pressure database to provide data support for subsequent maintenance and management, and facilitate trend analysis of cofferdam status and optimization of maintenance decisions. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0029] Figure 1 It is a framework flow chart of the present invention; Figure 2 It is a flow chart of cofferdam crack detection of the present invention; Figure 3 A schematic diagram of data set generation for water pressure prediction of cofferdams according to the present invention. DETAILED DESCRIPTION

[0030] To achieve the above objectives, the present invention is implemented through the following technical solutions: the present invention provides a method for monitoring the anti-seepage of cofferdams in tidal area cap construction, combined with Figure 1 , the method comprising: S1. Data Collection First, use several high-resolution cameras to regularly shoot the entire range of the cofferdam to obtain high-definition image data of the cofferdam surface with an image resolution of 1920*1080.

[0031] Second, water pressure sensors are installed on the cofferdam part below the water surface to monitor the water pressure changes outside the cofferdam in real time. Water pressure sensors are horizontally distributed on the cofferdam surface at fixed intervals to capture pressure information in different areas. At each time step t, water pressure data of the cofferdam is collected; in addition, local tidal data including tidal height, wave height and current speed are collected from the China Meteorological Network. The time step of tidal data is also t.

[0032] The timestamps of tidal data and cofferdam water pressure data are synchronized.

[0033] S2. Cofferdam status detection The collected high-definition image data is identified through computer vision and deep learning algorithms to determine whether the cofferdam has factors that may cause leakage, including cracks, posture deviations and deformations.

[0034] S21. Preprocess the high-definition image data through three methods: grayscale, smoothing filtering and histogram equalization.

[0035] Grayscale conversion is to convert a color image into a grayscale image to reduce the amount of calculation and highlight the edge information. Specifically, the three color channel (RGB) values ​​in the original image are averaged to obtain a grayscale value. The formula is:

[0036] in, Represents the grayscale value of the pixel in the i-th row and j-th column of the grayscale image. represents the R value of the i-th row and j-th column of the original high-definition image, represents the G value of the i-th row and j-th column of the original high-definition image, Represents the B value of the i-th row and j-th column of the original high-definition image.

[0037] Smoothing filtering is to remove random noise in the image and improve the accuracy of subsequent analysis. Specifically, the Gaussian function is used as a weight to calculate the weighted average of the pixel and its neighboring pixels. The shape of the Gaussian function is controlled by its standard deviation σ. For a given pixel position (i, j), its gray value SF_Gray(i, j) after smoothing filtering is calculated by the following formula:

[0038] Among them, W is the neighborhood window centered at (i, j), and the window size is k*l , is the value of the Gaussian function, defined as:

[0039] Histogram equalization is to enhance the contrast of the image by adjusting the histogram distribution of the image, making the cracks, deformation and other features more obvious. The mapping function of histogram equalization uses the cumulative distribution function, and the formula is:

[0040] Here L is the total number of gray levels, M and N are the length and width of the image respectively, is the number of pixels at the (i,j)th gray level after histogram equalization.

[0041] S22, crack detection 1) If Figure 2 As shown in the figure, the edge detection algorithm is used to detect cracks. The preprocessed image is convolved with the Sobel operator to highlight the areas where the grayscale changes significantly in the image, that is, the locations of the cracks. The specific steps are as follows: For each pixel in the image , respectively with G i and G j Perform convolution to obtain the horizontal gradient approximation Ii and the vertical gradient approximation I j .

[0042]

[0043] in Represents a two-dimensional convolution operation.

[0044] 2) Through I i and I j , the gradient magnitude and gradient direction of each pixel can be calculated. Gradient magnitude represents the strength of the edge, while the gradient direction Indicates the direction of the edge.

[0045]

[0046] 3) Perform non-maximum suppression on the gradient magnitude image M(i, j), specifically quantized to one of four main directions: horizontal, vertical, diagonal, and anti-diagonal. This is done by taking The angle difference with the four basic directions is realized, and the direction corresponding to the smallest difference is selected. The angles of the four basic directions are: horizontal direction: 0° or 180°; vertical direction: 90° or 270°; main diagonal direction: 45° or 225°; secondary diagonal direction: 135° or 315°.

[0047] According to the quantized direction, the nearest neighboring pixel in the gradient direction can be found. Specifically, When the direction is horizontal (0° or 180°), compare (i, j) with its left and right pixels (i+1,j)(i+1,j) and (i-1,j)(i-1,j).

[0048] When the direction is vertical (90° or 270°), compare (i, j) with the two pixels above and below it (i, j+1)(i, j+1) and (i, j-1)(i, j-1).

[0049] When the direction is diagonal (45° or 225°), compare (i, j) with its upper left and lower right pixels (i+1, j+1) and (i-1, j-1).

[0050] When the direction is anti-diagonal (135° or 315°), compare (i, j) with its upper right and lower left pixels (i+1, j-1) and (i-1, j+1).

[0051] If the gradient magnitude of pixel (i, j) is greater than that of its neighboring pixels, the gradient magnitude of that pixel is retained, otherwise it is set to 0.

[0052] 4) Perform threshold processing on the gradient magnitude image M(i,j), select a fixed threshold T, and regard all pixels with gradient magnitude greater than T as edge pixels, and other pixels as non-edge pixels.

[0053] Use the expansion operation to connect the fractured crack edges to form a complete crack outline. Specifically, select a rectangular structural element and align the origin of the structural element with For each pixel in the image, we determine whether the area covered by the structural element intersects with the edge pixels in the image. If there is an intersection, the pixel value of the pixel is set to the average value of all the covered edge pixels; if there is no intersection, the original value of the pixel is retained. The expanded image is represented as IF_Gray.

[0054] 5) Select a fixed threshold T b , convert IF_Gray into a binary image, that is, greater than T b The pixels are set to 255, otherwise they are set to 0.

[0055] Calculate the crack length: Use the skeleton extraction algorithm to reduce the lines of the binary image to a single pixel width, calculate the pixel length of the crack skeleton, and then convert the pixel length to the actual physical size based on the image resolution and the scale at which it was shot to obtain the crack length.

[0056] Calculate the crack width: Use the orthogonal skeleton line method to calculate the pixel width of the crack in the image, and then convert the pixel distance into the actual physical size according to the image resolution and the scale at the time of shooting to obtain the crack width.

[0057] S23, posture deviation detection Use the pre-trained CNN model to identify The model consists of three convolutional layers (the number of neurons is 32, 64 and 128 respectively, and the convolution kernel is 3*3), one pooling layer, one fully connected layer and one output layer. The number of neurons in the output layer is 6, representing the rotation angle (around the x, y and z axes) and translation (x, y and z directions) in three-dimensional space.

[0058] S24, deformation detection Use the pre-trained CNN model to identify The model consists of three convolutional layers (the number of neurons is 32, 64 and 128 respectively, and the convolution kernel is 3*3), one pooling layer, one fully connected layer and one classification layer. The number of neurons in the classification layer is 3, representing normal, convex and concave respectively.

[0059] S3. Water pressure prediction of cofferdam S31. Build a deep learning model. The main body of the model is the Bi-GRU structure. Bi-GRU can capture the temporal characteristics of data more comprehensively by considering both the forward information and the backward information in the time series. The forward GRU structure can be expressed as: Forward update gate:

[0060] in, is the sigmoid activation function, is the weight matrix of the update gate, is the hidden state at the previous time step, is the tide data of the current time step, is the bias term.

[0061] Forward reset gate:

[0062] in, is the weight matrix of the reset gate, is the bias term.

[0063] Forward to new candidate state:

[0064] in, is the weight matrix of the new candidate state, is the bias term, Represents element-wise multiplication.

[0065] Update the hidden state forward:

[0066] in, is the hidden state at the current time step.

[0067] The backward GRU structure is exactly the same as the forward one, and the final bidirectional GRU output is,

[0068] in, is the hidden state of the forward GRU, is the hidden state of the inverse GRU.

[0069] The Bi-GRU structure is followed by a fully connected layer to integrate the learned time series features.

[0070] S32: The input of the deep learning model is a fixed time window T s The time series of tide height, wave height, current speed and cofferdam water pressure are output as lag time A fixed time window T s This is to take into account the lag effect of tidal data on the cofferdam water pressure and the regularity of water pressure itself. The tidal data and cofferdam water pressure data at the current time step t are used to predict the lag time. The water pressure of the cofferdam, such as Figure 3 shown.

[0071] To find the best time window, select different time window sizes T s To verify, T s = {12t, 14t,16t, 18t, 20t, 22t}. To find the best lag time , select different lag times for verification, = {2t,4t,6t,8t,10t,12t}.

[0072] Using different time windows and lag times, the collected tidal data and cofferdam water pressure data are used to generate different data sets, which are divided into training and validation sets in a ratio of 8:2, respectively.

[0073] S33. Compare the mean square error (MSE) of the validation set of different data sets to determine the optimal time window and lag time. Input the time series of real-time tidal data and cofferdam water pressure data into the model to predict future Water pressure data of the cofferdam after time.

[0074] S4. Warning notification Set the crack length and width danger thresholds, rotation angle and translation danger thresholds. When the crack detection and posture deviation detection results exceed the set thresholds or the deformation detection results are convex and concave, the warning mechanism is immediately triggered to notify relevant personnel to carry out anti-seepage repair work on the cofferdam. The crack length danger threshold is set to 1 / 3 of the total length of the cofferdam, the crack width danger threshold is 0.5mm, the rotation angle danger threshold is 5°, and the translation danger threshold is 1 / 1000 of the cofferdam height.

[0075] A water pressure danger threshold for the cofferdam is set. When the predicted water pressure exceeds 70% of the cofferdam design limit, the early warning mechanism is immediately triggered and relevant personnel are notified to carry out anti-seepage reinforcement work on the cofferdam.

[0076] S5. Data storage and management A cofferdam status database is established to store the high-definition image data of the cofferdam and the cofferdam status detection results in the database, so that the historical status of the cofferdam can be checked in subsequent cofferdam management.

[0077] Establish a cofferdam water pressure database and store the water pressure sensor data and local tidal data as well as future cofferdam water pressure prediction results in the database. Regularly update the deep learning model parameters in S3 with the latest water pressure data and tidal data to ensure the accuracy of the cofferdam water pressure prediction.

[0078] The present invention improves the accuracy of the anti-seepage monitoring of the cofferdam. By combining high-definition images with water pressure sensor data, it not only monitors the changes in the external structure of the cofferdam, but also reflects the water pressure changes in the underwater part in real time, and comprehensively covers the key areas of the anti-seepage monitoring of the cofferdam. The image is processed and recognized by using computer vision technology and deep learning technology, which can efficiently and accurately detect cracks, deformations and posture deviations of the cofferdam, ensuring the early detection of anti-seepage problems.

[0079] The present invention introduces the Bi-GRU structural model and utilizes the forward and backward features of the time series to comprehensively capture the dynamic changes of tides and water pressure, significantly improve the accuracy of water pressure prediction, and regularly updates the deep learning model parameters based on the latest data to ensure that the prediction results always remain efficient and accurate.

[0080] When the detection result exceeds the set threshold, the present invention can automatically trigger an early warning and notify relevant personnel to take repair or reinforcement measures in time, reduce manual intervention, and save maintenance costs. The present invention constructs a comprehensive cofferdam status and water pressure database to provide data support for subsequent maintenance and management, and facilitate trend analysis of cofferdam status and optimization of maintenance decisions.

[0081] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for monitoring the anti-seepage of cofferdams for platform construction in tidal areas, characterized in that: The method includes: S1. Collect high-definition image data of the cofferdam surface, collect the time series of the cofferdam water pressure, and collect tidal data, including tidal height, wave height and current speed. The timestamps of tidal data and cofferdam water pressure data are synchronized; S2. Use computer vision and deep learning algorithms to identify the collected high-definition image data and determine whether there are factors that cause leakage in the cofferdam, including crack detection, posture deviation detection, and deformation detection; Detect the length and width of the cofferdam cracks, the rotation angle and translation of the cofferdam posture deviation, and whether the cofferdam is deformed; S3, build a deep learning model, use the collected tidal data and cofferdam water pressure data to predict the cofferdam water pressure in the future; S4. Setting dangerous thresholds for crack length and width, rotation angle and translation, and immediately triggering the warning mechanism when the crack detection and posture deviation detection results exceed the set thresholds or the deformation detection results are convex and concave; Set the water pressure danger threshold of the cofferdam to immediately trigger the early warning mechanism; S5. Establish a cofferdam status database, and store the high-definition image data of the cofferdam and the cofferdam status detection results in the database. A cofferdam water pressure database is established to store water pressure sensor data, local tidal data, and future cofferdam water pressure prediction results in the database, and the deep learning model parameters in S3 are regularly updated with the latest water pressure data and tidal data.

2. A method for monitoring the anti-seepage of cofferdams for platform construction in tidal areas according to claim 1, characterized in that: Before recognizing high-definition image data, the image data is preprocessed by grayscale, smoothing filtering and histogram equalization. The grayscale is as follows: Grayscale is to average the three color channel (RGB) values ​​in the original image to get a grayscale value. The formula is: , in, Represents the grayscale value of the pixel in the i-th row and j-th column of the grayscale image. represents the R value of the i-th row and j-th column of the original high-definition image, represents the G value of the i-th row and j-th column of the original high-definition image, Represents the B value of the i-th row and j-th column of the original high-definition image.

3. The anti-seepage monitoring method for cofferdam construction in tidal area according to claim 2, characterized in that: The smoothing filter uses a Gaussian function as a weight to calculate the weighted average of a pixel and its neighboring pixels. The shape of the Gaussian function is controlled by its standard deviation σ. For a given pixel position (i, j), the gray value SF_Gray(i, j) after smoothing filtering is calculated by the following formula: , Among them, W is the neighborhood window centered at (i, j), and the window size is k*l , is the value of the Gaussian function, defined as: 。 4. The anti-seepage monitoring method for cofferdam construction in tidal area according to claim 3, characterized in that: The histogram equalization is to enhance the contrast of the image by adjusting the histogram distribution of the image. The mapping function uses the cumulative distribution function, and the formula is: , Where L is the total number of gray levels, M and N are the length and width of the image respectively, is the number of pixels at the (i,j)th gray level after histogram equalization.

5. The anti-seepage monitoring method for cofferdams in tidal area cap construction according to claim 4 is characterized in that: The specific steps of crack detection are: 1) For the preprocessed image Each pixel in , respectively with G i and G j Perform convolution to obtain the horizontal gradient approximation I i and the vertical gradient approximation I j , the formula is as follows, , , in represents a two-dimensional convolution operation, G i and G j is the Sobel operator; 2) Calculate the gradient magnitude and gradient direction of each pixel, the gradient magnitude represents the strength of the edge, while the gradient direction Represents the direction of the edge, the formula is as follows, , ; 3) Perform non-maximum suppression on the gradient magnitude image M(i, j), specifically Quantize to one of the four main directions, namely horizontal, vertical, diagonal, and anti-diagonal. According to the quantized direction, the closest neighboring pixel in the gradient direction can be found. If the gradient magnitude of the pixel (i, j) is greater than the gradient magnitude of the neighboring pixel, the gradient magnitude of the pixel is retained, otherwise it is set to 0; 4) Perform threshold processing on the gradient magnitude image M(i,j), select a fixed threshold T, and regard all pixels with gradient magnitude greater than T as edge pixels, and other pixels as non-edge pixels; 5) Use the expansion operation to connect the fractured crack edges to form a complete crack outline. Specifically, select a rectangular structural element and align the origin of the structural element with For each pixel in the image, determine whether the area covered by the structural element intersects with the edge pixels in the image. If so, set the pixel value of the pixel to the average value of all covered edge pixels; if not, keep the original value of the pixel. The expanded image is represented as IF_Gray; 6) Select a fixed threshold T b , convert IF_Gray into a binary image, that is, greater than T b The pixel of is set to 255, otherwise it is set to 0; Use the skeleton extraction algorithm to reduce the lines of the binary image to a single pixel width, calculate the pixel length of the crack skeleton, and then convert the pixel length into the actual physical size according to the image resolution and the scale at the time of shooting to obtain the crack length; The pixel width of the crack in the image is calculated using the orthogonal skeleton method, and then the pixel distance is converted into the actual physical size according to the image resolution and the scale at the time of shooting to obtain the crack width.

6. The anti-seepage monitoring method for cofferdam construction in tidal area according to claim 4, characterized in that: The posture deviation detection is specifically to use the CNN model to identify the preprocessed image Image, the model consists of 3 convolutional layers, 1 pooling layer, 1 fully connected layer and 1 output layer in sequence. The number of neurons in the output layer is 6, representing the rotation angles of the cofferdam around the x, y and z axes in three-dimensional space and the translation in the x, y and z directions respectively.

7. The method for monitoring the anti-seepage of cofferdams for platform construction in tidal areas according to claim 4, characterized in that: The deformation detection is specifically to use the CNN model to identify the preprocessed image Image, the model consists of 3 convolutional layers, 1 pooling layer, 1 fully connected layer and 1 classification layer. The number of neurons in the classification layer is 3, representing normal, convex and concave cofferdams respectively.

8. The anti-seepage monitoring method for cofferdams in tidal area cap construction according to claim 1, characterized in that: The main body of the deep learning model in step S3 is a Bi-GRU structure, which includes a forward GRU structure and a backward GRU structure. The forward GRU structure can be expressed as: Forward update gate: , in, is the sigmoid activation function, is the weight matrix of the update gate, is the hidden state at the previous time step, is the tide data of the current time step, is the bias term, Forward reset gate: , in, is the weight matrix of the reset gate, is the bias term, Forward to new candidate state: , in, is the weight matrix of the new candidate state, is the bias term, represents element-wise multiplication; Update the hidden state forward: , in, is the hidden state at the current time step, The backward GRU structure is exactly the same as the forward one, and the final Bi-GRU output is, , in, is the hidden state of the forward GRU, is the hidden state of the inverse GRU, The Bi-GRU structure is followed by a fully connected layer.

9. The method for monitoring the anti-seepage of cofferdams for platform construction in tidal areas according to claim 1 or 8, characterized in that: The input to the deep learning model is a fixed time window T s The time series of tide height, wave height, current speed and cofferdam water pressure are output as lag time A fixed time window T s The water pressure time series, To find the best time window, select different time window sizes T s Verify to find the best lag time , select different lag times for verification, The mean square error (MSE) of the deep learning model under different time windows and different lag times was compared to determine the optimal time window and lag time, and the time series of real-time tidal data and cofferdam water pressure data were input into the model to predict future Water pressure data of the cofferdam after time.

Images