Tidal bore height measuring method based on oblique photography

Abstract

The invention discloses a tidal bore height measuring method based on oblique photography. According to the method, aiming at the problems of wide tidal bore range, strong power and difficulty in accurate observation, measurement of the tidal bore height is innovatively converted into identification and calculation of a tidal head edge curve by combining an unmanned aerial vehicle technology and an image processing method. The method comprises the following steps: synchronously tracking a tidal bore by using an unmanned aerial vehicle, and capturing a tidal head image; based on the brightness difference between the tide head and the background, edge curves of the tide head top surface and the tide front horizontal plane are determined through an edge detection method; calculating the pixel coordinate difference of the curve in the image; and finally, combining the flight height of the unmanned aerial vehicle and camera parameters, and calculating the actual height of the tidal bore according to the pixel coordinate difference of the edge curves of the tidal bore top surface and the tidal front horizontal plane in the image. The method provided by the invention has the characteristics of safety, continuous observation capability and suitability for wide water areas, and provides an innovative tidal bore height measurement solution.

Description

Technical Field

[0001] The present invention belongs to the field of smart water conservancy, and relates to a tidal bore observation method, and specifically to a tidal bore height measurement method based on oblique photography. Background Art

[0002] Tidal bores are a unique and spectacular natural phenomenon that occurs in estuaries where rivers meet the ocean. The phenomenon is characterized by a rapid rise in tidal water, often accompanied by strong flow rates and water level changes. Tidal bores have been recorded in about 450 estuaries or bays around the world, which usually have specific geographical and hydrological conditions that allow tidal water to converge and produce a strong upwelling effect. At high tide, the seawater rises due to tidal forces. When this seawater flows into the estuary, if it encounters high river flows and narrow river channels, it will form a highly concentrated water flow, which in turn forms high waves in the estuary area, and sometimes even a spectacular water wall phenomenon. China's famous Qiantang River tidal bore can be called the "world's largest tidal bore", attracting a large number of tourists and researchers to observe it every year. In addition to the Qiantang River, tidal bores also occur in Aojiang River, Lingjiang River and the northern branch of the modern Yangtze River estuary. The huge energy of tidal bores not only has a significant impact on estuary geology and sediment movement, but also on estuary ecosystems and local shipping activities. The rapid and unpredictable nature of tidal bores makes observation and research challenging.

[0003] A tidal bore is a strong tidal wave caused by tidal forces, characterized by a sudden rise in water levels and an accelerated flow rate, which forms a crest and causes rolling waves on the water surface, usually called a tidal head. The key indicators in tidal bore research include tidal bore height, maximum tidal bore flow rate, tidal bore propagation speed, tidal bore steepness and tidal bore period. The tidal bore height is a key parameter to describe the physical characteristics of a tidal bore, which refers to the vertical distance of the highest point of the tidal bore head relative to the average water level. This height is an indicator of the significance of tidal bore water level changes and reflects the scale and energy of the tidal bore. Under the same other conditions, the greater the tidal bore height, the stronger the tidal wave force of the tidal bore, and the greater the impact on rivers, coastlines and related waters. Therefore, the tidal bore height is an important basis for assessing the potential impact of tidal bores and conducting tidal bore management.

[0004] At present, the methods for studying the height of tidal bore can be mainly divided into two types: field measurement and theoretical analysis. For example, Xie Dongfeng et al. used a water level recorder to monitor the water level changes in the tidal bore process in real time in the paper "Research on the Hydrodynamic Characteristics of Qiantang River Tidal Bore Based on Measured Data" (Hydrodynamic Research and Progress, 2012) to directly calculate the height of the tidal bore. The observed height of the tidal bore usually ranges from tens of centimeters to several meters. Qi Lan et al. used turbulence models and theoretical analysis of hydrodynamic characteristics in the paper "CFD Numerical Simulation of Tidal Bore Flow" (Journal of Water Resources and Hydropower Engineering, 2019) to simulate the height of the tidal bore under different conditions.

[0005] The shortcomings of the current methods for studying tidal bore heights mainly include: for field measurement methods, it is necessary to install measuring equipment such as water level gauges in the tidal bore area. Due to the strong power and unpredictability of tidal bores, these devices are extremely difficult to install and are easily damaged. In addition, maintenance and data collection face great challenges under extreme weather conditions. In addition, field measurements are usually limited to specific locations and it is difficult to reflect the changes in tidal bore heights in a wider area. As for numerical simulation methods, although it can provide the spatial distribution and time series of tidal bore heights, the accuracy of the model is limited by the accuracy of the description of the physical process used and the setting of boundary conditions. Numerical models often cannot reach the ideal state, which may cause the simulated tidal bore height to be inconsistent with the actual situation.

[0006] In view of the above problems, the present invention proposes a new method for measuring the height of the tidal bore using a drone assisted by oblique photography technology. When the tidal bore moves forward with strong power, the water flow will hit the water surface. Such violent movement will produce a large amount of foam, which makes the tidal head area and the background area clearly different. Utilizing the particularity of the tidal head area in the image, the top surface of the tidal bore head and the horizontal plane before the tide in the image are detected with the help of image edge detection. The height of the tidal bore is calculated by the position difference between the top surface of the tidal bore head and the horizontal plane before the tide in the image, combined with the height and angle of the drone shooting. Summary of the invention

[0007] The purpose of the present invention is to make up for the shortcomings of the existing tidal height measurement method. By combining drones and image processing methods, a tidal height measurement method based on oblique photography is proposed. The calculation problem of the tidal height is innovatively converted into the calculation of the height of the tidal head edge line in the image. This method has the advantages of high safety, dynamic real-time monitoring and good data richness.

[0008] A method for measuring tidal height based on oblique photography, characterized in that it comprises the following steps:

[0009] 1) The multi-rotor drone is equipped with a camera to keep tracking and shooting synchronously with the tidal bore in the river section where the tidal bore is observed, and the drone is kept to continuously capture the tidal bore image A from the front of the tidal bore. The flight altitude of the drone and the camera parameters are determined according to the tidal bore height, observation range and weather conditions of the tidal bore to be observed;

[0010] 2) Grayscale conversion is performed on the tidal bore image A taken by the drone, the color information of the image is converted into grayscale levels, and the image is denoised using Gaussian filtering technology to obtain the tidal bore image B, in preparation for further image analysis;

[0011] 3) Apply edge detection algorithm to identify and extract the edge curve L1 of the top surface of the tidal head and the edge curve L2 of the horizontal surface before the tide;

[0012] 4) Perform morphological dilation operations on the detected edge curves L1 and L2 to enhance the features in the image and connect the existing broken edges;

[0013] 5) Perform morphological opening operation to further remove small noise points in the image and optimize the integrity and clarity of edge curves L1 and L2;

[0014] 6) After edge enhancement and noise removal, the edge curve c1 of the top surface of the tidal head and the edge curve c2 of the horizontal surface before the tide are clearly drawn in the image;

[0015] 7) calculating the difference in pixel coordinates between the tidal head top surface edge curve c1 and the pre-tidal horizontal surface edge curve c2 in the image according to the pixel positions of the tidal edge curves c1 and c2 in the image;

[0016] 8) According to the flight altitude of the UAV and the camera shooting parameters, the actual height of the tidal bore is calculated by combining the pixel coordinate difference between the edge curve c1 of the top surface of the tidal bore head and the edge curve c2 of the horizontal surface before the tide in the tidal bore image.

[0017] The specific steps in step 1 are: first, a gimbal camera is mounted on the drone, and the gimbal self-stabilization feature can ensure the stability of the camera's visual axis, and the image is obtained by tilt photography; then the drone keeps dynamic synchronous tracking with the tidal bore, as shown in FIG. Figure 2 and Figure 3 As shown, the camera is made to continuously capture images of the tidal bore from the front of the tidal bore to obtain tidal bore image A.

[0018] The method for denoising in step 2 comprises the following steps: 2-1: Convert the image to grayscale, converting the original color image into a grayscale image; 2-2: Gaussian filtering. The processed image usually loses some details and reduces noise. 2-3: Obtain a grayscale tidal bore image B including the tidal bore head. The step 2 processing is beneficial for reducing image noise and subsequent image processing (such as edge detection) because it can reduce the number of false edges that the algorithm may identify.

[0019] In step 3, after edge extraction, an image of the same size as the original image will be generated, in which the edges are represented by white pixels and the non-edge areas are black. This produces a clear contour image that highlights the structural boundaries of the tidal head in the image.

[0020] Step 4 is a morphological image processing method, which is used to enlarge the bright area in the image. The specific steps are as follows: 4-1: Create a structural element k1, which is used as a kernel in the morphological operation. Align the center of the structural element to each pixel of the image. This structural element will determine the range and shape of the dilation operation. 4-2: In the area covered by the structural element k1, the pixel values ​​in the image are compared with the corresponding positions of the structural element; 4-3: If the pixel value of at least one position in the structural element k1 is higher than a certain threshold in the grayscale image, then the value of the pixel corresponding to the center of the structural element is set to the maximum value. This process can be vividly understood as that the influence range of each pixel is expanded to the shape and size defined by the structural element k1, thereby causing the expansion of the object in the image.

[0021] The step 5 is a morphological image opening operation processing method, which is used to reduce the bright area in the image. The specific steps are as follows: the morphological opening operation includes corrosion and expansion operations, and a structural element k2 is defined. The corrosion operation aligns the structural element k2 with each pixel of the image and defines the central pixel value as the minimum value within the coverage range of the structural element, thereby purposefully reducing the size of the object, removing small interferences, and making the tide head curve more refined. The expansion operation is just the opposite. It keeps the alignment of the structural element k2 with the pixel and expands the central pixel value to the maximum value within the coverage range, thereby partially restoring the tide head curve that has been reduced in the corrosion step, but will not restore the small noise points that have been removed. The opening operation optimizes the image by the strategy of first corrosion and then expansion, effectively smoothing the edges, eliminating small protrusions, and maintaining the basic structure of the tide head curve unchanged. The effect of this operation is adjusted with the setting of the size of the structural element k2. The larger the structural element, the more prominent the processing effect.

[0022] The method for drawing edge curves of the top surface of the tidal head and the front tidal horizontal surface in step 6 is specifically as follows: 6-1: Find edge curves in the morphologically processed black and white image and find the edge of the white foreground object; 6-2: Ignore the inner edge curves and keep the outer edge curves among all edge curves. Also remove the shorter redundant edge curves and keep the longer edge curves. Ensure that there are only two edge curves in the vertical direction, one of which represents the top surface of the tidal head and the other represents the horizontal surface before the tide. The edge curve of the top surface of the tidal head is represented by c1, and the edge curve of the horizontal surface before the tide is represented by c2; 6-3: Create a black image of the same size and type as the processing image, and draw the edge curves c1 and c2 in white on a black background.

[0023] The specific calculation method of calculating the coordinate position difference between the edge curve of the top surface of the tidal head and the edge curve of the horizontal surface before the tide in the image in step 7 is: 7-1: Traverse the width of the image, that is, all possible values ​​of the x coordinate. If there are edge curve points above and below this x coordinate position, calculate the average y coordinate of these points, and find the pixel coordinate difference of the two edge curves in the y direction at this x position, that is, the vertical distance between the pixels, in terms of the number of pixels between them; 7-2: Calculate the average value of the pixel coordinate difference in the y direction between the edge curve c1 of the tidal head top surface and the edge curve c2 of the tidal front horizontal surface

[0024] The specific method of step 8 is: 8-1: Calculate the pixel size δ of the drone camera. The pixel is the smallest unit of each picture, that is, each small pixel grid in an image. Each pixel grid is a square, so it can be understood that the length and width of the pixel are the same. 8-2: If Figure 3 As shown, according to the vertical image GSD calculation formula GSD=δ*h / f combined with the rotation angle of the main optical axis of the oblique camera, the approximate resolution near the center point of the oblique image (the actual physical size of a pixel) can be obtained by the following formula. Where δ is the size of a single pixel of the sensor; H is the relative altitude of the UAV; f is the focal length of the camera; and α is the angle between the camera optical axis and the horizontal plane. 8-3: Calculate the tidal bore height h. In the formula, It is the average value of the pixel coordinate difference between the edge curve of the top surface of the tide head and the edge curve of the horizontal surface before the tide in the y direction; GSD is the actual physical size corresponding to a pixel in the image.

[0025] The beneficial effects of the present invention are:

[0026] The present invention converts the calculation problem of the tidal bore height into the calculation of the height of the tidal head edge line in the image, and can indirectly calculate the tidal bore height.

[0027] The present invention does not require the deployment of an observation platform in the tidal river section, and does not require the placement of measuring instruments and equipment such as water level and flow velocity on the platform. It is a non-contact observation method with high observation safety and no interference with the development of tidal bores, which is extremely important for the observation of tidal bores with strong power and great destructive power.

[0028] The present invention uses a UAV as an observation platform, giving full play to the advantages of the UAV's mobility and flexibility, so that the observation of the tidal height is more convenient and safer.

[0029] The UAV always keeps synchronous flight with the tidal bore head, so the invention can carry out dynamic and continuous observation of the tidal bore height.

[0030] The drone always maintains a certain flight altitude and has a larger photographing field of view, so the present invention can cover and calculate the tidal height of a larger range of waters. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 This is a flow chart of the tidal height measurement method based on oblique photography.

[0032] Figure 2 This is an image of the tidal bore taken on-site by a drone.

[0033] Figure 3 This is a schematic diagram of a tidal bore captured by an oblique drone.

[0034] Figure 4 It is a schematic diagram of the process of drawing the edge curves of the top surface of the tidal head and the horizontal surface before the tide. DETAILED DESCRIPTION

[0035] The present invention is further explained below in conjunction with the accompanying drawings and specific embodiments.

[0036] Step 1: Use a multi-rotor drone equipped with a gimbal camera to capture a sequence of tidal bore images. The self-stabilizing gimbal is used to isolate the changes in the drone's posture, body vibration, and wind resistance disturbance, to ensure the stability of the drone's onboard camera's visual axis, and to enable the camera to continuously capture tidal bore images from the front. Figure 2 and Figure 3 As shown in the figure, the multi-rotor UAV keeps dynamic synchronous tracking with the tidal bore in the tidal bore observation section, and always keeps the UAV in front of the tidal bore head. The flight altitude of the UAV and the camera parameters are calculated and determined according to the specific requirements of the tidal bore observation. The focal length f of the UAV camera used in this experiment is 24mm, the flight altitude H of the UAV is 40m, and the angle α between the camera optical axis and the horizontal plane is 30°.

[0037] Step 2: Perform grayscale conversion and Gaussian filtering on the acquired image: (1) Grayscale conversion In image recognition, the human eye can easily distinguish the edges and boundaries of different objects in the image. For computers, it is more complicated to directly process color information. The color of a color image is stored in the format of RGB, i.e., red, green, and blue color space. After conversion to a grayscale image, the image will only contain grayscale information and remove the color information, which is beneficial for computers to process tidal images. Specifically, the process of converting a color image to a grayscale image involves the transformation of the color space, in which the values ​​of the three color channels of red R, green G, and blue B of each color pixel are converted into a single grayscale value. This process can be expressed by a mathematical formula, Y′=0.299R+0.587G+0.114B In the formula, Y' represents the calculated grayscale value, and R, G, and B are the values ​​of the red, green, and blue channels of the original color pixel, respectively. This formula is based on the characteristics of the human visual system, and different weights are given according to the different sensitivities of the human eye to different colors. The green color has the highest weight because the human eye is most sensitive to green, and the blue color has the lowest weight because the human eye is least sensitive to blue. Grayscale conversion calculation process: 1) For each pixel, read the values ​​of its R, G, and B channels. These values ​​are generally in the range of 0 to 255; 2) Substitute the values ​​of the R, G, and B channels into the above conversion formula; 3) The calculated Y' value is the corresponding grayscale value. This grayscale value reflects the brightness information of the pixel; 4) Repeat steps 1 to 3 for all pixels in the image. (2) Gaussian filtering Noise is often introduced during the image capture process. Based on the Gaussian function, the image is smoothed by weighted averaging of the surrounding pixels. The weight is determined by the Gaussian function, and the closer the point is to the center pixel, the higher the weight. This method can reduce noise while retaining the overall structure of the image. The mathematical expression of the two-dimensional Gaussian function is: Here, x and y represent the horizontal and vertical distances of the pixel from the center of the kernel (the center pixel); σ is the standard deviation of the Gaussian distribution, which is calculated from the size of the Gaussian kernel. The processing process of Gaussian filtering is as follows: 1) For each pixel in the image, locate its 3x3 neighborhood; 2) Multiply each pixel value in this neighborhood by the corresponding Gaussian kernel weight; 3) Calculate the sum of these products; 4) Assign this sum to the center pixel. Using Gaussian filtering to smooth an image can effectively reduce image noise while retaining the main features of the image, such as edges, to a certain extent. Because Gaussian filtering takes into account the spatial relationship between a pixel and its surrounding pixels, it provides a more natural and smoother effect than simple mean filtering (the new value of each pixel is a simple average of the values ​​of its neighboring pixels). The results after grayscale conversion and Gaussian filtering are as follows Figure 4 (a) shown.

[0038] Step 3: Apply edge detection algorithm to identify and extract edge curves of the top surface of the tidal head and the horizontal surface before the tide. Edge detection is a method used to identify points in an image where there are significant changes in brightness, usually corresponding to object boundaries. Edge detection algorithms attempt to determine the location and degree of changes in pixel brightness, which represent changes in structure in the image. The process usually involves mathematical gradient calculations, because gradients can describe the strength and direction of changes in image brightness. The gradient is a measure of brightness change and can be expressed as: Where f is the brightness function of the image, Gx and Gy are the brightness gradient components of the image in the horizontal and vertical directions respectively. These components can be approximated by applying an edge detection filter (such as the Sobel filter). The specific process of the Sobel filter includes: 3-1: Gradient calculation: Set the Sobel operator (convolution kernel) in the horizontal and vertical directions: For each pixel in the image, these two operators are used to perform convolution operations with the pixel and its surrounding pixels respectively to calculate the gradient components of the point in the horizontal and vertical directions. 3-2: Gradient magnitude and direction: Combine the two calculated gradient components to calculate the gradient magnitude and direction of each pixel: Where G is the gradient magnitude and θ is the gradient direction. 3-3: Non-maximum suppression: Used to refine the gradient image. For each pixel in the gradient image, it is compared with its two neighboring points along the gradient direction (i.e., the two neighboring points on the gradient line). If the current pixel is not the maximum value, it is suppressed (set to 0), so that only the most obvious edges are retained. 3-4: Dual Threshold and Edge Tracking: Two thresholds are used to distinguish strong edges, weak edges, and non-edges. The high threshold is used to detect strong edges, and the low threshold is used to distinguish weak edges from non-edges. For pixels with gradient magnitudes lower than the low threshold, they are directly considered not to be edges. For pixels with gradient magnitudes higher than the high threshold, they are considered to be strong edges. For values ​​between the two, if they are connected to a strong edge, they are retained as edges; otherwise, they are suppressed. Through the above processing, the edge detection algorithm can highlight the edge curve of the tidal bore in the image, providing a basis for further image processing. Figure 4 (b) as shown.

[0039] Step 4: Morphological processing: Use dilation operation to connect broken edges and filter out noise. Dilation operation is used to connect broken edges and filter out noise. It can increase the size of edge curves in the image and strengthen the connected parts, so that the broken edges of the edge curves are connected together, while reducing small, isolated noise points. The dilation operation is based on the convolution operation of the structural element (or kernel) and the image. 4-1: The structure element is a predefined matrix that determines the way of dilation. In the present invention, the structure element k1 is defined as a 2x1 rectangle, which means that the dilation process will increase the connection of pixels in the horizontal direction and align the center of the structure element to each pixel of the image; 4-2: In the area covered by the structural element, compare the pixel values ​​in the image with the corresponding positions of the structural element; 4-3: If the pixel value of at least one position in the structural element k1 is higher than a certain threshold in the grayscale image, then the value of the target pixel (the pixel corresponding to the center of the structural element) is set to the maximum value. This process can be figuratively understood as the "influence range" of each pixel is expanded to the shape and size defined by the structural element, causing the curve in the image to expand horizontally. In this tidal image processing, the dilation operation was performed three times. Multiple dilations can further enhance the degree of connection between objects in the image and be more effective. The experimental results are as follows Figure 4 (c) as shown.

[0040] Step 5: Morphological processing: Opening operation removes small noise. Opening operation is a morphological operation, which is a combination of erosion operation followed by a dilation operation. This operation is particularly useful for removing small noise points while keeping the main structure of the image unchanged. Opening operation can effectively remove small protruding parts of an object while having little effect on the overall size and shape of the object. The opening operation object is the image that has been processed by dilation before, and the structure element k2 used is a 2x1 rectangle. The opening operation mainly has two steps: 5-1: Erosion operation, the purpose of erosion is to reduce the size of the object in the image. The erosion operation can be regarded as a minimum convolution operation between the image and the structural element: first, align the center of the structural element k2 with each pixel of the image, and then compare the pixel value in the image with the corresponding position of the structural element k2 in the area covered by the structural element k2, and then set the value of the pixel corresponding to the center of the structural element k2 to the minimum value in the area. This allows small objects in the image, i.e. noise, to be removed, while large objects become more detailed. 5-2: Dilation operation,Following the dilation operation, the dilation operation increases the size of the object in the image. The dilation operation can be regarded as a maximum convolution operation between the image and the structural element k2: Similarly, the center of the structural element k2 is aligned with each pixel of the image, and within the area covered by the structural element k2, the pixel value in the image is compared with the corresponding position of the structural element k2, and the value of the target pixel is set to the maximum value in the area. In this way, large objects that were previously reduced in the erosion step will be "restored" to close to their original size, while small objects, i.e. noise, that were removed by the erosion step will not be restored in the dilation step because they have already been removed by the erosion operation. Through this series of operations, the opening operation effectively removes small noise in the image while keeping the main structure of the target basically unchanged. Figure 4 (d) as shown.

[0041] Step 6: involves drawing the edge curve c1 of the top surface of the tidal head and the edge curve c2 of the front tidal surface in the processed black and white image. First, by finding the outer edge of the white foreground object, ignoring the inner edges in the image and removing the shorter redundant edge curves, only the longer curves are retained to ensure that two edge curves are clearly identified in the vertical direction: one represents the top surface of the tidal head and the other represents the front tidal surface. Then, on a newly created black background image, these two edge curves c1 and c2 are accurately drawn in white to clearly show the key features of the tidal bore.

[0042] Step 7: Based on the position of the tidal edge curve provided in step 6, calculate the difference in pixel coordinates between the tidal head top edge curve c1 and the pre-tidal horizontal edge curve c2 in the image. This process is done by traversing the width of the image, that is, every pixel point on the x-axis, looking for points on the upper and lower lines that match the x-coordinates of these pixel points, and then calculating the y-coordinate difference (i.e., vertical distance) between the upper and lower points. Store these vertical distances (y-differences) in an array. Finally, calculate the average value of the y-difference array. This average value represents the average vertical distance between the edge curve c1 of the tidal head and the edge curve c2 of the tidal front horizontal surface in the image, which is a quantitative representation of the tidal height. The value is about 461 pixels.

[0043] Step 8: In this experiment, the flying altitude H of the UAV is 40m, and the angle α between the camera optical axis and the horizontal plane is 30°. The focal length of the drone camera used is 24mm. The pixel size is calculated by the following formula, and δ is 1.5um. According to the following ground sampling distance GSD calculation formula, it can be concluded that GSD near the center point of the oblique image = 0.5cm / pixel, which means that the actual size of one pixel is 0.5cm*0.5cm. Finally, the tidal bore height is calculated Pixel*0.5cm=230.5cm, that is, the height of the tidal bore is about 2.3m.

[0044] The above are preferred embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art who is familiar with the technical field shall, within the technical scope disclosed by the present invention, make any changes or substitutions according to the technical solution of the present invention, which shall be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. A method for measuring tidal height based on oblique photography, characterized in that: The following steps are involved: Step 1: The multi-rotor drone is equipped with a camera to keep tracking and shooting synchronously with the tidal bore in the river section where the tidal bore is observed, and the drone is kept to continuously capture the tidal bore image A from the front of the tidal bore. The flight altitude of the drone and the camera parameters are determined according to the tidal bore height, observation range and weather conditions of the tidal bore to be observed; Step 2: grayscale conversion is performed on the tidal bore image A taken by the drone, the color information of the image is converted into grayscale levels, and the image is denoised by Gaussian filtering technology to obtain the tidal bore image B, in preparation for further image analysis; Step 3: Apply edge detection algorithm to identify and extract the edge curve L1 of the top surface of the tidal head and the edge curve L2 of the horizontal surface before the tide; Step 4: Perform morphological dilation operations on the detected edge curves L1 and L2 to enhance the features in the image and connect the existing broken edges; Step 5: Perform morphological opening operation to further remove small noise points in the image and optimize the integrity and clarity of edge curves L1 and L2; Step 6: After edge enhancement and noise removal, the edge curve c1 of the top surface of the tidal head and the edge curve c2 of the horizontal surface before the tide are clearly drawn in the image; Step 7: According to the pixel positions of the tidal edge curves c1 and c2 in the image, the difference in pixel coordinates between the tidal head top surface edge curve c1 and the pre-tidal horizontal surface edge curve c2 in the image is calculated; Step 8: According to the flight altitude of the drone and the camera shooting parameters, the actual height of the tidal bore is calculated by combining the pixel coordinate difference between the edge curve c1 of the top surface of the tidal bore head and the edge curve c2 of the horizontal surface before the tide in the tidal bore image.

2. The method for measuring tidal height based on oblique photography according to claim 1, characterized in that: The specific steps of step 1 are: using a multi-rotor drone equipped with a gimbal camera to shoot tidal bore image A, wherein the self-stabilizing gimbal is used to isolate the changes in the drone's posture and the vibration and wind resistance disturbance of the drone body to ensure the stability of the visual axis of the drone's onboard camera. The multi-rotor drone maintains dynamic synchronous tracking with the tidal bore in the tidal bore observation river section, and always keeps the drone in front of the tidal bore head, so that the camera can continuously capture tidal bore images from the front.

3. The method for measuring tidal height based on oblique photography according to claim 1, characterized in that: The grayscale conversion and denoising method in step 2 comprises the following steps: 2-1: Convert the image to grayscale, converting the original color image into a grayscale image; 2-2: Gaussian filtering. The processed image usually loses some details and reduces noise. 2-3: Obtain a grayscale tidal bore image B including the tidal bore head.

4. The method for measuring tidal height based on oblique photography according to claim 1, characterized in that: The specific steps of edge detection in step 3 are: the brightness of the target is an easily distinguishable feature, and the places where the brightness changes significantly in the image are found, which usually correspond to the boundaries of the target. The following is a brief description of the process: 3-1: Gradient calculation, using the Sobel operator, applying convolution operation to each pixel of the image, and calculating the brightness gradient components Gx and Gy in the horizontal and vertical directions respectively; 3-2: Gradient amplitude and direction, combining the horizontal gradient component Gx and the vertical gradient component Gy to calculate the gradient amplitude G and gradient direction θ of each pixel to determine the strength and direction of the edge; 3-3: Non-maximum suppression: Non-maximum suppression is performed on the gradient image to refine the edges in the gradient image and ensure that only the most obvious edge lines are retained; 3-4: Double threshold and edge tracking. By setting two thresholds to distinguish strong edges from weak edges, points with gradient amplitudes lower than the low threshold are regarded as non-edges, and those higher than the high threshold are regarded as strong edges. Points in between are decided whether to be retained as edges based on their connectivity with the strong edge. This series of steps highlights the tidal edge curves L1 and L2 in the image, providing a basis for further image analysis.

5. The method for measuring tidal height based on oblique photography according to claim 1, characterized in that: The specific steps of the morphological dilation operation in step 4 are as follows: the dilation operation is based on the convolution operation of the structural element and the image, and the process of the dilation operation is as follows: 4-1: First, define a structure element k1 and align the center of the structure element to each pixel of the image; 4-2: In the area covered by the structural element k1, compare the pixel values ​​in the image with the corresponding positions of the structural element; 4-3: If the pixel value of at least one position in the structural element k1 is higher than a certain threshold in the grayscale image, then the value of the pixel corresponding to the center of the structural element is set to the maximum value. This process can be figuratively understood as that the influence range of each pixel is expanded to the shape and size defined by the structural element k1, thereby causing the expansion of the object in the image.

6. The method for measuring tidal height based on oblique photography according to claim 1, characterized in that: The specific steps of the morphological opening operation in step 5 are as follows: the morphological opening operation includes corrosion and expansion operations, and a structural element k2 is defined. The corrosion operation aligns the structural element k2 with each pixel of the image and defines the central pixel value as the minimum value within the coverage range of the structural element, thereby purposefully reducing the object size, removing small interferences, and making the tidal head curve more refined. The expansion operation is just the opposite. The structural element k2 is aligned with the pixel and the central pixel value is expanded to the maximum value within the coverage range, thereby partially restoring the tidal head curve that has been reduced in the corrosion step, but the small noise points that have been removed will not be restored. The opening operation optimizes the image by the strategy of first corrosion and then expansion, effectively smoothing the edges, eliminating small protrusions, and maintaining the basic structure of the tidal head curve unchanged. The effect of this operation is adjusted with the setting of the size of the structural element k2. The larger the structural element, the more prominent the processing effect.

7. The method for measuring tidal height based on oblique photography according to claim 1, characterized in that: The specific steps of drawing the tidal head edge curve in step 6 are: searching for edge curves in the black and white image after morphological processing, finding the edge of the white foreground object, searching and identifying the outer edge curve of the white foreground object in the image, ignoring the inner edge and removing the shorter edge curves, and retaining only the longer curves, so that only two main edge curves are identified in the vertical direction, representing the top surface of the tidal head and the horizontal plane before the tide, respectively; then, the edge curve c1 of the top surface of the tidal head and the edge curve c2 of the horizontal plane before the tide are drawn in white on a newly created black background image.

8. The method for measuring tidal height based on oblique photography according to claim 1, characterized in that: The process of calculating the coordinate position difference of the two edge curves c1 and c2 in the image in step 7 is as follows: first, traverse the width of the image, and for each x-coordinate position, if there are edge curve points above and below the position, calculate the y-coordinate difference of these points, so as to obtain the pixel coordinate difference of the two edge curves in the y direction at the x-coordinate position, that is, the pixel distance in the vertical direction, and then summarize the y-direction pixel coordinate differences of all x-coordinate positions, and calculate their average value as the average vertical distance between the edge curve c1 on the top surface of the tide head and the edge curve c2 on the horizontal surface before the tide 9. The method for measuring tidal height based on oblique photography according to claim 1, characterized in that: The step of calculating the tidal bore height in step 8 is as follows: based on the sensor size of the drone camera and the pixel size of the image, the size δ of the camera single pixel, that is, the physical size of each pixel grid, is calculated. Then, the calculation formula of the ground sampling distance GSD is used, combined with the relative height H of the drone flight, the focal length f of the camera, and the angle α between the main optical axis of the camera and the horizontal plane, to calculate the actual physical resolution near the center point of the inclined image. Finally, the average value of the pixel coordinate difference in the y direction between the edge curve c1 of the top surface of the tidal bore head and the edge curve c2 of the horizontal surface before the tide is calculated. This average value is multiplied by the previously calculated GSD value to obtain the actual tidal height h. This method relies on the accurate calculation of the actual physical size of the pixels and the distance between the edge curves of the tidal head in the image, effectively using drone images to quantify the tidal height.

Images