Mudflat beach face elevation monitoring method and system based on mudflat photovoltaic video image

Abstract

The invention provides a mudflat beach face elevation monitoring method and system based on a mudflat photovoltaic video image, and the method comprises the steps: collecting a mudflat video at regular time through a plurality of cameras installed below a photovoltaic panel on a photovoltaic support, carrying out the frame extraction of the mudflat video, and obtaining a mudflat image key frame; elevation information of the beach face of the mud flat is extracted through the key frame of the mud flat image, and the elevation information of the beach face of the mud flat is subjected to three-dimensional modeling to generate a three-dimensional elevation model of the mud flat; and setting observation lines and observation points in the three-dimensional elevation model, and regularly calculating and recording elevation data of the observation points. According to the method, visual overall form reconstruction can be provided through the generated three-dimensional model covering the mudflat terrain, including the characteristics of the tidal creek, the sand dam, the slope surface and the like, and the monitoring efficiency and accuracy of the mudflat change can be effectively improved.

Description

Technical Field

[0001] This application belongs to the technical field of monitoring the elevation change of tidal flat surfaces, and relates to an elevation monitoring method, particularly to a method and system for monitoring the elevation of tidal flat surfaces based on photovoltaic video images of tidal flats. Background Art

[0002] Tidal flats are an important part of the coastline. Their ecological systems and geomorphic features are affected by natural factors such as tides, ocean currents, and wind waves, resulting in frequent siltation and erosion phenomena. Traditional methods for monitoring the elevation of tidal flats mainly rely on manual observations and regular measurements, which are inefficient and difficult to obtain real-time data. With the development of the tidal flat photovoltaic industry, more and more photovoltaic power stations have been completed in the tidal flat waters along the coast of China. This will inevitably change the sedimentation state under the natural state of the tidal flats, and may increase the trend of siltation or erosion of the pile foundations on the tidal flats, posing potential safety hazards to engineering construction. In order to monitor the elevation change of the tidal flat surface in real time, photovoltaic facilities can be combined with monitoring technologies, and high-definition cameras installed on the photovoltaic panel supports can be used for real-time monitoring, which can effectively improve the monitoring efficiency and accuracy of tidal flat changes.

[0003] The technical development of tidal flat elevation observation has gradually progressed from primitive to modern, from small-scale to large-scale, from single to multi-means, and from low-efficiency to high-efficiency. Traditional observation methods include contact and non-contact methods, which are observed from different angles of points, lines, and surfaces. The representative "marker pile method" in point observation technology requires firmly inserting a benchmark into the tidal flat sediments, and then measuring the elevation of the tidal flat surface by measuring the length of the exposed part. This method will damage the natural condition of the tidal flat and also change the microenvironment of erosion and deposition. Only one point can be measured with one pile, and the efficiency is low. Point observations are not convenient to be densely carried out on the tidal flat. If we want to reflect the complete erosion and deposition change trend on the cross-section line of the tidal flat facing the sea, line observation technology is more often used.

[0004] Furthermore, it can be known that line observation technology commonly uses technologies such as RTK. Generally, a reference station is set up by selecting control points near the shore, and the mobile station needs to manually "run the beach" to mark points according to the set survey line. The efficiency is also low and real-time observation is not possible. Due to the limitations of the observation range and working efficiency, the point and line observation methods of tidal flat erosion and deposition mainly target limited specific points or key research areas.

[0005] To comprehensively understand the erosion and deposition situation of the tidal flat, it is more appropriate to select diverse planar observation technologies. Planar observation technologies include: airborne (spaceborne) laser measurement, synthetic aperture radar interferometry, unmanned aerial vehicle oblique photogrammetry, remote sensing feature line technology, hyperspectral inversion, etc. These methods also have their own advantages, disadvantages, and applicable scenarios. Obviously, in the scenario where a large number of photovoltaic panels block the view above the tidal flat photovoltaic field area, these methods for monitoring the elevation of the tidal flat are not applicable.

[0006] The purpose of the present invention is to design a method for monitoring the elevation of the tidal flat surface based on the video images of tidal flat photovoltaic, aiming at the problems existing in the above-mentioned prior art. Summary of the Invention

[0007] The purpose of the present application is to provide a method and system for monitoring the elevation of the tidal flat surface based on the video images of tidal flat photovoltaic, which are used to solve the problems in the prior art that due to the excessive dependence on manual work, the elevation information of the tidal flat surface cannot be monitored in real time and dynamically, and the accuracy of the elevation monitoring of the key change areas of the tidal flat is too low.

[0008] In the first aspect, the present application provides a method for monitoring the elevation of the tidal flat surface based on the video images of tidal flat photovoltaic, including the following steps: obtaining the tidal flat video of the target area to be monitored; performing frame extraction processing on the tidal flat video to obtain the key frames of the tidal flat images; obtaining the elevation information of the tidal flat surface according to the key frames of the tidal flat images; performing three-dimensional modeling based on the elevation information of the tidal flat surface to generate a three-dimensional elevation model of the tidal flat; setting observation lines and observation points in the three-dimensional elevation model of the tidal flat, and regularly calculating and recording the elevation data of the observation points.

[0009] In one implementation manner of the first aspect, performing frame extraction processing on the tidal flat video to obtain the key frames of the tidal flat images includes: reading the tidal flat video frame by frame, preprocessing each frame of the tidal flat video image to obtain the first tidal flat video image; constructing a scene classification prediction model based on the tidal flat video; inputting the tidal flat video image into the scene classification prediction model to obtain the scene type corresponding to the tidal flat video image; if the current tidal flat video image is a tidal flat exposure scene, marking the frame corresponding to this tidal flat exposure scene as a candidate key frame, and calculating the proportion of the changed pixel points of each frame of the candidate key frame; if the proportion of the changed pixel points is greater than the change threshold, marking this frame as the key frame of the tidal flat image, otherwise skipping the current frame.

[0010] In one implementation manner of the first aspect, constructing a scene classification prediction model based on the tidal flat video includes: extracting a number of tidal flat video image frames based on the tidal flat video; performing marking on the tidal flat video image frames to obtain scene labels; the scene labels include any one or a combination of high tide period, low tide period, tidal flat exposure, and tidal flat flooding; constructing a scene classification prediction model through a convolutional neural network, using the image frames as the input and the scene labels as the output, and training the scene classification prediction model.

[0011] In an implementation of the first aspect, calculating the proportion of changed pixel points in each candidate key frame includes: calculating the grayscale value difference of corresponding pixels in two adjacent frames before and after; obtaining changed pixel points according to a filtering rule and calculating the proportion of the changed pixel points; the filtering rule includes: setting a grayscale value difference threshold; filtering out the changed pixel points when the grayscale value difference is less than the grayscale value difference threshold; and counting other changed pixel points when the grayscale value difference is greater than or equal to the grayscale value difference threshold; the calculation formula for the grayscale value difference is:

[0012] difference = I t (x, y) - I t-1 (x, y)

[0013] where, I t (x, y) represents the pixel value of the current frame; I t-1 (x, y) represents the pixel value of the previous frame.

[0014] In an implementation of the first aspect, obtaining the elevation information of the tidal flat surface according to the key frames of the tidal flat images includes: obtaining a number of image features based on the key frames of the tidal flat images, and matching the image features of the key frames of the tidal flat images under different probes in the same tidal flat area; setting a similarity threshold; selecting feature points with a similarity greater than the similarity threshold from the number of image features, and calculating the horizontal parallax of each feature point in its tidal flat image; generating a parallax map through the horizontal parallax; obtaining the depth information of the parallax map by using the triangulation method, and converting the depth information into elevation information; the calculation formula for the horizontal parallax is:

[0015] d = x1 - x2

[0016] where, d represents the horizontal parallax value; x1 and x2 respectively represent the horizontal coordinates of the feature points in two tidal flat images.

[0017] In an implementation of the first aspect, generating a parallax map through the horizontal parallax; obtaining the depth information of the parallax map by using the triangulation method, and converting the depth information into elevation information includes: calculating the depth information of the parallax map based on the parallax information; the parallax information includes: parallax value, probe focal length, baseline distance between two probes, probe depth value; calibrating the probe depth value with the local elevation reference of the tidal flat to obtain the elevation information of the tidal flat surface; the calculation formula for the probe depth value is:

[0018]

[0019] where, Z represents the probe depth value; f represents the probe focal length; B represents the baseline distance between two probes; d represents the parallax value.

[0020] In an implementation of the first aspect, the three-dimensional elevation model of the tidal flat is as follows:

[0021]

[0022] Z = Z

[0023] where (X, Y, Z) represents the three-dimensional coordinates of the pixel point; Z represents the depth information; (u, v) represents the pixel point coordinates; (c x , c y ) represents the coordinates of the principal point of the image calibrated by the camera.

[0024] In an implementation of the first aspect, observation lines and observation points are set in the three-dimensional elevation model of the tidal flat, and the elevation data of the observation points are regularly calculated and recorded, including: setting a strip of observation lines along the slope direction in the three-dimensional elevation model of the tidal flat; uniformly setting a number of observation points on each of the observation lines; in the three-dimensional elevation model of the tidal flat, obtaining the three-dimensional coordinates closest to the coordinates of the observation point, and obtaining the elevation data of the observation point through the closest three-dimensional coordinates.

[0025] In an implementation of the first aspect, in the three-dimensional elevation model of the tidal flat, obtaining the three-dimensional coordinates closest to the coordinates of the observation point and obtaining the elevation data of the observation point through the closest three-dimensional coordinates includes: calculating the elevation gradient between adjacent observation points on the same observation line; setting a gradient threshold; when the elevation gradient exceeds the gradient threshold, the interval between the observation points on this observation line is changed to a second interval, and the second interval is smaller than the first interval.

[0026] In the second aspect, the present application provides a tidal flat elevation monitoring system based on tidal flat photovoltaic video images, including: an acquisition module for acquiring the tidal flat video of the target area to be monitored; a frame extraction module for performing frame extraction on the tidal flat video to obtain key frames of the tidal flat images; an elevation information acquisition module for obtaining the tidal flat elevation information according to the key frames of the tidal flat images; a model construction module for performing three-dimensional modeling based on the tidal flat elevation information to generate a three-dimensional elevation model of the tidal flat; an elevation monitoring module for setting observation lines and observation points in the three-dimensional elevation model of the tidal flat and regularly calculating and recording the elevation data of the observation points.

[0027] As described above, the tidal flat elevation monitoring method and system based on tidal flat photovoltaic video images provided by the present application have the following beneficial effects:

[0028] (1) The method for monitoring the elevation of the tidal flat surface based on the video images of the tidal flat photovoltaic provided by this application selects appropriate positions in the already built offshore tidal flat photovoltaic field area. By installing cameras on the photovoltaic brackets, taking advantage of the characteristics of their fixed positions, the photovoltaic brackets can provide relatively high installation positions, and the field of view of the cameras can cover a relatively large area of the tidal flat, enabling unattended automated data collection and reducing the frequency and intensity of manual surveys. During the ebb and flow of the tide or the process of erosion and deposition, the change of the image information in the video has time criticality. By means of scene recognition and frame extraction based on the change of pixel points, the key states of these dynamic change nodes can be extracted.

[0029] (2) In this application, by calculating the depth information, elevation feature points or image parallax of the pixel points in the image, the plane image data can be converted into the elevation data of the actual tidal flat terrain. Using the extracted elevation data, a three-dimensional model covering the tidal flat terrain can be generated, including tidal channels, sandbars, and slope features, providing an intuitive reconstruction of the overall morphology.

[0030] (3) In this application, by setting observation lines (along the slope direction) and observation points (refined to key areas) in the three-dimensional elevation model, the elevation monitoring of the key change areas of the tidal flat becomes more systematic and accurate. By regularly calculating and recording the elevation values of the observation points, dynamic comparative analysis can be carried out on the processes of tidal flat deposition, erosion, regional expansion or retreat.

[0031] (4) This application has wide applicability. Description of the Drawings

[0032] Figure 1 It shows a schematic diagram of the application scenario of the method for monitoring the elevation of the tidal flat surface based on the video images of the tidal flat photovoltaic described in this application in an embodiment.

[0033] Figure 2 It shows a schematic flow chart of the method for monitoring the elevation of the tidal flat surface based on the video images of the tidal flat photovoltaic described in this application in an embodiment.

[0034] Figure 3 It shows a schematic diagram of the installation of the camera on the photovoltaic bracket in an embodiment of the method for monitoring the elevation of the tidal flat surface based on the video images of the tidal flat photovoltaic described in this application.

[0035] Figure 4 It shows a schematic flow chart of S2 in the method for monitoring the elevation of the tidal flat surface based on the video images of the tidal flat photovoltaic described in this application.

[0036] Figure 5 It shows a schematic flow chart of S3 in the method for monitoring the elevation of the tidal flat surface based on the video images of the tidal flat photovoltaic described in this application.

[0037] Figure 6It shows a schematic flowchart of S5 in an embodiment of the method for monitoring the elevation of the tidal flat surface based on the video images of the tidal flat photovoltaic in the present application.

[0038] Figure 7 It shows a schematic diagram of setting an observation line on the tidal flat in an embodiment of the method for monitoring the elevation of the tidal flat surface based on the video images of the tidal flat photovoltaic in the present application.

[0039] Figure 8 It shows a schematic diagram of the principle structure in an embodiment of the system for monitoring the elevation of the tidal flat surface based on the video images of the tidal flat photovoltaic in the present application.

[0040] Figure 9 It shows a schematic diagram of the principle structure in an embodiment of the device for monitoring the elevation of the tidal flat surface based on the video images of the tidal flat photovoltaic in the present application.

[0041] Description of component labels

[0042] 11 Video acquisition module

[0043] 12 Frame extraction module

[0044] 13 Information conversion module

[0045] 14 3D modeling module

[0046] 15 Observation and calculation module

[0047] 81 Acquisition module

[0048] 82 Frame extraction module

[0049] 83 Elevation information acquisition module

[0050] 84 Model construction module

[0051] 85 Elevation monitoring module

[0052] 91 Processor

[0053] 92 Memory Detailed implementation manners

[0054] The following uses specific specific examples to illustrate the implementation manners of the present application. Those skilled in the art can easily understand other advantages and effects of the present application from the content disclosed in this specification. The present application can also be implemented or applied through other different specific implementation manners, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present application. It should be noted that, without conflict, the following embodiments and the features in the embodiments can be combined with each other.

[0055] It should be noted that the illustrations provided in the following embodiments only schematically illustrate the basic concept of the present application. Therefore, only the components related to the present application are shown in the drawings, rather than being drawn according to the number, shape, and size of the components in actual implementation. The type, quantity, and ratio of each component in actual implementation can be arbitrarily changed, and the component layout type may also be more complex.

[0056] Please refer to Figure 1 , which shows a schematic diagram of an application scenario of the method for monitoring the elevation of the tidal flat surface based on the photovoltaic video image of the tidal flat in an embodiment of the present application.

[0057] The following embodiments of the present application provide a method and system for monitoring the elevation of the tidal flat surface based on the photovoltaic video image of the tidal flat. In the video acquisition module 11, several cameras are installed below the photovoltaic panels on the photovoltaic brackets to regularly acquire tidal flat videos; the frame extraction module 12 performs frame extraction on the tidal flat videos to obtain key frames of the tidal flat images; the elevation information of the tidal flat surface is extracted from the key frames of the tidal flat images in the information conversion module 13, and the elevation information of the tidal flat surface is used to generate a three-dimensional elevation model of the tidal flat through three-dimensional modeling by the three-dimensional modeling module 14; observation lines and observation points are set in the three-dimensional elevation model in the observation and calculation module 15, and the elevation data of the observation points are regularly calculated and recorded. The present application solves the problems in the prior art that due to excessive reliance on manual work, it is impossible to monitor the elevation information of the tidal flat surface in real time and dynamically, and the accuracy of monitoring the elevation of the key change areas of the tidal flat is too low. The present application takes pictures from the height of the photovoltaic panel, can cover a large range of tidal flat areas, and realizes planar monitoring; performs unattended automated data acquisition, reduces the frequency and intensity of manual surveys; through a three-dimensional model covering the tidal flat terrain, it can provide an intuitive overall shape reconstruction; at the same time, makes the elevation monitoring of the key change areas of the tidal flat more systematic and accurate.

[0058] Next, the method and system for monitoring the elevation of the tidal flat surface based on the photovoltaic video image of the tidal flat provided in the embodiments of the present application will be described in detail with reference to the accompanying drawings in the embodiments of the present application.

[0059] Please refer to Figure 2 and Figure 3 , which respectively show a schematic flow diagram of the method for monitoring the elevation of the tidal flat surface based on the photovoltaic video image of the tidal flat in an embodiment of the present application and a schematic installation diagram of the camera on the photovoltaic bracket in an embodiment of the method for monitoring the elevation of the tidal flat surface based on the photovoltaic video image of the tidal flat of the present application. As Figure 2 and Figure 3 shown, this embodiment provides a method for monitoring the elevation of the tidal flat surface based on the photovoltaic video image of the tidal flat.

[0060] The method for monitoring the elevation of the tidal flat surface based on the tidal flat photovoltaic video image specifically includes the following steps:

[0061] S1. Obtain the tidal flat video of the target area to be monitored.

[0062] In this embodiment, according to the research requirements, several probes (such as cameras) installed below the photovoltaic panels on the photovoltaic brackets are used to collect the tidal flat video at regular intervals.

[0063] Specifically, the shooting angle of the adopted camera is perpendicular to the tidal flat slope; there is a certain baseline distance between the cameras in the same tidal flat area.

[0064] In this step, the construction of the tidal flat photovoltaic field area provides convenience in terms of installation, power, safety, etc. for the on-site layout of the video monitoring system. This method makes full use of these convenient conditions to carry out the monitoring work on the elevation change of the tidal flat surface.

[0065] Please continue to refer to Figure 3 .

[0066] For example: As Figure 3 can be seen, select a suitable location in the already built offshore tidal flat photovoltaic field area, and fixedly install a high-definition camera on the photovoltaic panel bracket to collect the video images of the tidal flat surface. To ensure that the video screen can cover the required observation area, at least two or more cameras need to be installed in the same area. The shooting angle of the camera should be as perpendicular to the tidal flat slope as possible, and the shooting directions of the two cameras should be located at both ends or both sides of the shooting area respectively, with a certain baseline distance, so as to improve the comprehensiveness of the monitoring. Connect the video signals of the cameras to the hard disk video recorder and the computer host, set the timing recording function, and regularly (such as every hour or every half hour) collect the video images of the tidal flat surface to capture the influence of tidal changes and other dynamic factors on the tidal flat.

[0067] S2. Perform frame extraction processing on the tidal flat video to obtain the key frames of the tidal flat images. Please refer to Figure 4 , which shows the schematic flow chart of S2 in the method for monitoring the elevation of the tidal flat surface based on the tidal flat photovoltaic video image described in this application. As Figure 4 shown, the S2 includes the following steps:

[0068] According to the above, even if the video has been collected at regular intervals, each frame of the original video data often contains a large amount of redundant information, and the content in consecutive frames may not change much. Therefore, the key frame extraction method is adopted here. The main purpose of key frame extraction is to reduce redundancy and retain the core information that can reflect the changes in the elevation of the tidal flat. The environment of the tidal flat may change slowly for a long time (such as the gradual rise and fall of the tide, etc.), and there may be no significant difference in consecutive regularly collected frames; directly using all frames will increase the computational and storage burden and have limited contribution to data analysis. Therefore, through key frame extraction, frames without changes (such as the tidal flat with different illumination angles but unchanged surface features) can be filtered out. Retain the frames that reflect the key changes of the tidal flat (such as significant changes in the tide level, the tidal flat being completely exposed or submerged).

[0069] S21, read the tidal flat video frame by frame, preprocess each frame of the tidal flat video image (such as denoising and de-graying, etc.) to obtain the preprocessed tidal flat video image, that is: the first tidal flat video image;

[0070] S22, construct a scene classification prediction model based on the tidal flat video.

[0071] In this embodiment, several frames of the tidal flat video image are extracted based on the tidal flat video; the frames of the tidal flat video image are marked to obtain scene labels; the scene labels include: any one or a combination of high tide period, low tide period, tidal flat exposure, and tidal flat inundation; a scene classification prediction model is constructed through a convolutional neural network, taking the image frame as the input and the scene label as the output, and training the scene classification prediction model.

[0072] S23, input the tidal flat video image into the scene classification prediction model to obtain the scene type corresponding to the tidal flat video image;

[0073] S24, if the current tidal flat video image is a tidal flat exposure scene, mark the frame corresponding to this tidal flat exposure scene as a candidate key frame, and calculate the proportion of the changed pixel points in each frame of the candidate key frame. It includes: calculating the gray value difference between the corresponding pixels of two adjacent frames before and after; obtaining the changed pixel points according to the filtering rule and calculating the proportion of the changed pixel points; the filtering rule includes: setting a gray value difference threshold; when the gray value difference of the changed pixel points is less than the gray value difference threshold, filter them; when the gray value difference of the changed pixel points is greater than or equal to the gray value difference threshold, then count the other changed pixel points.

[0074] The calculation formula for the gray value difference is:

[0075] difference = I t (x,y) - I t-1 (x,y)

[0076] where, It (x, y) represents the pixel value of the current frame; I t-1 (x, y) represents the pixel value of the previous frame.

[0077] If the proportion of changed pixel points is greater than the change threshold, mark this frame as a key frame of the tidal flat image; otherwise, skip the current frame.

[0078] In this embodiment, first, read the tidal flat video frame by frame, perform denoising and grayscale conversion on each frame of the tidal flat video image to obtain the preprocessed tidal flat video image; input the tidal flat video image into the scene classification prediction model to obtain the scene type corresponding to the tidal flat video image.

[0079] Furthermore, the scene classification prediction model can be trained through the following steps:

[0080] Specifically, extract a large number of image frames from the tidal flat monitoring video as the input data for training the scene classification prediction model. Each frame corresponds to a scene label, and the scene labels include but are not limited to: high tide period, low tide period, tidal flat exposure, tidal flat flooding, etc.; construct a scene classification prediction model through a convolutional neural network, use the image frames as the input and the scene labels as the output to train the scene classification prediction model.

[0081] In this step, the tidal flat exposure scene (such as: the surface of the tidal flat is exposed after ebb tide) usually reflects the most significant features of the tidal flat landform and the dynamic changes of the ecosystem, and is high-value data for collecting key elevation information and conducting tidal flat modeling. The flooded scene (such as: the underwater state of the tidal flat or being completely covered by the tide) is not conducive to extracting landform features, so it does not need to be used as a key frame. If frames are directly extracted based on feature changes, redundant frames may be extracted due to unnecessary changes in the environment (such as: light, waves or background). Scene classification priority can directly filter out non-target scenes that do not belong to "tidal flat exposure", thus more efficiently screening out data that meets the monitoring requirements. After being trained, the scene classification model can accurately label tidal flat scenes (such as: high tide, low tide, exposure, flooding). Therefore, in the case of defining frame criticality, useful frames can be directly and quickly selected according to the scene labels, without having to calculate the differences for all frames, saving computing resources.

[0082] In this embodiment, if the current tidal flat video image is a tidal flat exposure scene, mark this frame as a candidate key frame and calculate the proportion of changed pixel points in each frame of the candidate key frames.

[0083] Specifically, first, calculate the gray value difference of the corresponding pixels in two consecutive frames. The calculation formula is:

[0084] difference = I t (x, y) - I t-1 (x, y)

[0085] Among them, I t (x, y) represents the pixel value of the current frame; I t-1 (x, y) represents the pixel value of the previous frame.

[0086] Then, filter out the changing pixel points with the gray value difference less than the gray value difference threshold, count the other changing pixel points, and calculate the proportion of the changing pixel points.

[0087] In this embodiment, if the proportion of the changing pixel points is greater than the change threshold, mark this frame as the key frame of the tidal flat image, otherwise skip the current frame.

[0088] Specifically, scene classification is relatively accurate in the stage of annotating "tidal flat exposure", but the tidal flat exposure scene still contains a large number of frames, and not all frames have significant changes in the tidal flat. By further evaluating the feature changes, the precise screening of frames can be completed. Focus on the significant changes in the tidal flat exposure scene through the gray value difference, and extract the information frames that generate dynamic evolution during the exposure process (such as: the tide recedes, more tidal flats are exposed, and the surface is turned over). Reduce the redundant frame extraction within the continuous segment of the tidal flat exposure scene (i.e., the redundant pixel changes in the continuous similar scenes).

[0089] Therefore, "feature change" is a supplementary discriminant condition, and frames with new significance can be screened out from the continuous frames of the exposure scene. Scene classification is given priority to, and frames related to the task (such as: tidal flat exposure) can be quickly filtered out. Further use the feature change value to screen in the target scene to control the number of frames, thereby improving the frame extraction efficiency.

[0090] S3. Obtain the tidal flat elevation information according to the key frames of the tidal flat images. Please refer to Figure 5 , which shows the flow schematic diagram of S3 in the tidal flat elevation monitoring method based on tidal flat photovoltaic video images described in this application. As Figure 5 shown, the S3 includes the following steps:

[0091] S31. Obtain a number of image features based on the key frames of the tidal flat images, and match the image features of the key frames of the tidal flat images under different probes in the same tidal flat area;

[0092] S32. Set a similarity threshold;

[0093] S33. Select the feature points with similarity greater than the similarity threshold from a number of the image features, and calculate the horizontal parallax of each feature point in its tidal flat image;

[0094] S34. Generate a parallax map through the horizontal parallax;

[0095] S35. Obtain the depth information of the parallax map by using triangulation and convert the depth information into elevation information, including: calculating the depth information of the parallax map based on the parallax information, where the parallax information includes parallax value, probe focal length, baseline distance between two probes, and probe depth value; calibrating the probe depth value with the local elevation datum of the tidal flat to obtain the elevation information of the tidal flat surface.

[0096] In this embodiment, first, extract the image features of the key frames of the tidal flat images and match the image features of the key frames of the tidal flat images obtained by different cameras in the same tidal flat area.

[0097] Then, select the feature points with a similarity greater than the similarity threshold and calculate the horizontal parallax of each feature point in its image. The calculation formula is as follows:

[0098] d = x1 - x2

[0099] where d represents the horizontal parallax value; x1 and x2 respectively represent the horizontal coordinates of the feature points in the two tidal flat images.

[0100] Next, generate a parallax map through the calculation of the horizontal parallax, calculate the depth information of the parallax map through triangulation, and convert the depth information into elevation information.

[0101] Specifically, calculate the depth information of the parallax map through the parallax d, the focal length f of the camera, the baseline distance B between the two cameras, and the depth value Z of the camera. The calculation formula is as follows:

[0102]

[0103] where Z represents the probe depth value; f represents the probe focal length; B represents the baseline distance between two probes; d represents the parallax value.

[0104] Calibrate the depth value Z with the local elevation datum of the tidal flat to obtain the elevation information of the tidal flat surface.

[0105] As can be seen from the above, the key frames of the tidal flat images contain the two-dimensional visible data at a specific moment of the tidal flat. By calculating the relative position relationship between the images (such as parallax) and the information of camera calibration, we can deduce the depth information (height) of the tidal flat surface according to the geometric relationship. Since different cameras are in different positions (or the image frames are collected from different viewpoints during the movement of the camera), the same feature (such as a point on the tidal flat surface) in the key frames will have a relative displacement in the camera view. This displacement (parallax) provides the distance information about the feature in the depth direction (Z-axis).

[0106] Feature point detection and matching are carried out between the key frames of the tidal flat, which can clarify the projection positions (i.e., mapping relationships) of the same three-dimensional points in the scene in different images, thereby providing a basis for restoring depth. The SIFT (SIFT is an image feature detection and description algorithm that can extract key points with scale-invariance and rotation-invariance and their descriptors from images) or SURF (SURF is an optimized feature detection and description algorithm. It is inspired by SIFT, but significantly improves the running speed through mathematical optimization and fast calculation methods while maintaining robustness to scale changes and rotation changes) algorithm is used to extract key feature points. This algorithm is prior art and will not be elaborated in this invention.

[0107] Although the depth Z can be calculated through geometric derivation, the extracted depth value needs to be calibrated with the local elevation datum of the tidal flat area (such as the data of the 1985 National Elevation Datum). Due to tides, erosion, sedimentation, etc., there will be slight height differences on the tidal flat surface. The changes of these elevation features in multi-view images can be captured by parallax and depth measurement and used to generate elevation information.

[0108] S4, based on the elevation information of the tidal flat surface, perform three-dimensional modeling to generate a three-dimensional elevation model of the tidal flat.

[0109] In this embodiment, based on the depth information Z, the three-dimensional coordinates (X, Y, Z) of each pixel point are calculated to generate a three-dimensional elevation model.

[0110] The three-dimensional elevation model of the tidal flat is:

[0111]

[0112]

[0113] Z = Z

[0114] where, (X, Y, Z) represents the three-dimensional coordinates of the pixel point; Z represents the depth information; (u, v) represents the pixel point coordinates; (c x , c y ) represents the coordinates of the principal point of the image calibrated by the camera.

[0115] In this step, since the terrain of the tidal flat area is determined by height, slope changes, and surface features. One-dimensional or two-dimensional elevation data can only qualitatively describe elevation changes and cannot comprehensively and quantitatively analyze the tidal flat surface. And one point of elevation data can show the height of a certain area, but it cannot reveal how the surface undulates. However, the three-dimensional elevation model can construct the specific shape of the tidal flat area and provide rich surface details.

[0116] S5. Set observation lines and observation points in the three-dimensional elevation model of the tidal flat, and regularly calculate and record the elevation data of the observation points. Please refer to Figure 6 , which shows a schematic flowchart of S5 in the tidal flat elevation monitoring method based on tidal flat photovoltaic video images described in this application. As Figure 6 shown, S5 includes the following steps:

[0117] S51. Set an observation line along the slope direction in the three-dimensional elevation model of the tidal flat;

[0118] S52. Uniformly set a number of observation points on each of the observation lines;

[0119] S53. In the three-dimensional elevation model of the tidal flat, obtain the three-dimensional coordinates closest to the coordinates of the observation point, and obtain the elevation data of the observation point through the closest three-dimensional coordinates. It includes: calculating the elevation gradient between adjacent observation points on the same observation line; setting a gradient threshold; when the elevation gradient exceeds the gradient threshold, change the interval between the observation points on this observation line to a second interval, and the second interval is smaller than the first interval.

[0120] In this embodiment, multiple observation lines along the slope direction are set on the three-dimensional elevation model, and a number of observation points are uniformly set on each observation line at the first interval; in the three-dimensional elevation model, find the three-dimensional coordinates closest to the coordinates of the observation point, and obtain the elevation data of the observation point through the closest three-dimensional coordinates.

[0121] Further, in the three-dimensional elevation model, finding the three-dimensional coordinates closest to the coordinates of the observation point and obtaining the elevation data of the observation point through the closest three-dimensional coordinates are further performed: calculating the elevation gradient between adjacent observation points on the same observation line, and if the elevation gradient exceeds the gradient threshold, change the interval between the observation points on this observation line to a second interval, and the second interval is smaller than the first interval.

[0122] In this step, as Figure 7 shown, in the tidal flat environment, the elevation change usually has a direction (such as: deposition and erosion along the slope direction). Setting observation lines can systematically distribute monitoring points and capture the overall characteristics of the tidal flat area. When setting observation lines along the slope, the changes of the scouring area and the deposition area over time can be visually monitored. During the flood tide period, the height of the damaged slope area decreases, and during the ebb tide period, the deposition zone expands or the surface rises. The observation points arranged in a specific direction allow the analysis of the continuity of the elevation change and obtain the spatial law of the tidal flat erosion and deposition process. By uniformly arranging observation points on the observation line, the number and distribution of observation points can be optimized according to the research objectives and the topographic characteristics of the tidal flat, improving the monitoring efficiency.

[0123] Since the coordinate points in the three-dimensional model are generated by discrete sampling or gridding, the coordinates of the theoretical observation points may not exactly correspond to the model data. By finding the nearest points, the discreteness and precision limitations of the model can be accommodated, interpolation errors can be avoided, and the flexibility and robustness of the data can be enhanced.

[0124] Too many distributed points will waste computing resources, and uneven distribution may miss key areas. In some areas with large variations (such as tidal creeks, toe of slope, etc.), by arranging dense observation points, these key features can be captured. The gradient can be calculated through elevation to analyze the change of terrain slope. The areas with large slope changes can be found in advance, and during the next observation, the number of observation points can be increased to ensure the accuracy of observation.

[0125] Furthermore, based on the daily, weekly, and monthly data, analyze the elevation changes of the observation lines and observation points. Use statistical analysis methods (such as linear regression, time series analysis, etc.) to evaluate the trend of beach erosion or sedimentation.

[0126] The method for monitoring the elevation of the beach surface based on the photovoltaic video images of the tidal flat provided by this application selects a suitable location in the already built offshore tidal flat photovoltaic field area. By installing a camera on the photovoltaic support, the characteristics of its fixed position can be utilized. The photovoltaic support can provide a relatively high installation position, and the field of view of the camera can cover a relatively large area of the tidal flat, enabling unattended automated data collection and reducing the frequency and intensity of manual surveys. During the ebb and flow of the tide or the process of erosion and deposition, the change of the image information in the video has time-criticality. Through scene recognition and frame extraction based on the change of pixel points, the key states of these dynamic change nodes can be extracted. In this application, by calculating the depth information, elevation feature points or image parallax of the pixel points in the image, the planar image data can be converted into the elevation data of the actual tidal flat terrain. Using the extracted elevation data, a three-dimensional model covering the tidal flat terrain can be generated, including tidal creeks, sandbars, and slope features, providing an intuitive reconstruction of the overall morphology. At the same time, by setting observation lines (along the slope direction) and observation points (refined to key areas) in the three-dimensional elevation model, the elevation monitoring of the key change areas of the tidal flat becomes more systematic and accurate. By regularly calculating and recording the elevation values of the observation points, a dynamic comparative analysis of the processes of tidal flat sedimentation, erosion, regional expansion or retreat can be carried out.

[0127] The protection scope of the method for monitoring the elevation of the beach surface based on the photovoltaic video images of the tidal flat described in the embodiments of this application is not limited to the execution order of the steps listed in this embodiment. Any scheme achieved by adding or reducing steps of the prior art and replacing steps according to the principle of this application is included in the protection scope of this application.

[0128] The embodiment of the present application further provides a monitoring system for the elevation of a tidal flat surface based on tidal flat photovoltaic video images. The monitoring system for the elevation of a tidal flat surface based on tidal flat photovoltaic video images can implement the method for monitoring the elevation of a tidal flat surface based on tidal flat photovoltaic video images described in the present application. However, the implementation device of the method for monitoring the elevation of a tidal flat surface based on tidal flat photovoltaic video images described in the present application includes, but is not limited to, the structure of the monitoring system for the elevation of a tidal flat surface based on tidal flat photovoltaic video images listed in this embodiment. Any structural deformation and replacement of the prior art made according to the principle of the present application are included in the protection scope of the present application.

[0129] The following will describe in detail the monitoring system for the elevation of a tidal flat surface based on tidal flat photovoltaic video images provided in this embodiment with reference to the drawings.

[0130] This embodiment provides a monitoring system for the elevation of a tidal flat surface based on tidal flat photovoltaic video images, including:

[0131] Please refer to Figure 8 , which shows the schematic diagram of the principle structure of the monitoring system for the elevation of a tidal flat surface based on tidal flat photovoltaic video images described in the present application in an embodiment. As Figure 8 shown, the monitoring system for the elevation of a tidal flat surface based on tidal flat photovoltaic video images includes: an acquisition module 81, a frame extraction module 82, an elevation information acquisition module 83, a model construction module 84, and an elevation monitoring module 85.

[0132] The acquisition module 81 is used to acquire the tidal flat video of the target area to be monitored.

[0133] In this embodiment, according to the research requirements, a number of probes (such as cameras, etc.) installed under the photovoltaic panels on the photovoltaic brackets are used to collect tidal flat videos regularly.

[0134] Specifically, the shooting angle of the camera used is perpendicular to the tidal flat slope; there is a certain baseline distance between the cameras in the same tidal flat area.

[0135] In this step, the construction of the tidal flat photovoltaic field provides convenience in terms of installation, power, safety, etc. for the on-site layout of the video monitoring system. This method makes full use of these convenient conditions to carry out the monitoring work on the elevation change of the tidal flat surface.

[0136] The frame extraction module 82 is connected to the acquisition module 81 and is used to perform frame extraction processing on the tidal flat video to obtain the key frames of the tidal flat images.

[0137] In this embodiment, the tidal flat video is read frame by frame, and each frame of the tidal flat video image is preprocessed (such as denoising and de - graying, etc.) to obtain the preprocessed tidal flat video image, that is: the first tidal flat video image; a scene classification prediction model is constructed based on the tidal flat video; the tidal flat video image is input into the scene classification prediction model to obtain the scene type corresponding to the tidal flat video image; if the current tidal flat video image is a tidal flat exposure scene, the frame corresponding to this tidal flat exposure scene is marked as a candidate key frame, and the proportion of the changed pixel points of each frame of the candidate key frame is calculated.

[0138] Specifically, calculate the gray - value difference of the corresponding pixels of two adjacent frames before and after; obtain the changed pixel points according to the filtering rule, and calculate the proportion of the changed pixel points; the filtering rule includes: setting a gray - value difference threshold; when the changed pixel points with a gray - value difference less than the gray - value difference threshold are filtered; when the changed pixel points with a gray - value difference greater than or equal to the gray - value difference threshold, then count the other changed pixel points. If the proportion of the changed pixel points is greater than the change threshold, mark this frame as the key frame of the tidal flat image, otherwise skip the current frame.

[0139] The elevation information acquisition module 83 is used to obtain the tidal flat beach elevation information according to the key frame of the tidal flat image.

[0140] In this embodiment, a number of image features are obtained based on the key frame of the tidal flat image, and the image features of the key frames of the tidal flat images under different probes in the same tidal flat area are matched; a similarity threshold is set; the feature points with a similarity greater than the similarity threshold are selected from the number of image features, and the horizontal parallax of each feature point in its tidal flat image is calculated; a parallax map is generated through the horizontal parallax; the depth information of the parallax map is obtained by using the triangulation method, and the depth information is converted into elevation information. It includes: calculating the depth information of the parallax map based on the parallax information; the parallax information includes: parallax value, probe focal length, baseline distance between two probes, probe depth value; calibrate the probe depth value with the local elevation reference of the tidal flat to obtain the elevation information of the tidal flat beach surface.

[0141] The model construction module 84 is used to perform three - dimensional modeling based on the tidal flat beach elevation information to generate a three - dimensional elevation model of the tidal flat.

[0142] In this embodiment, based on the depth information Z, the three - dimensional coordinates (X, Y, Z) of each pixel point are calculated to generate a three - dimensional elevation model.

[0143] The elevation monitoring module 85 is used to set observation lines and observation points in the three - dimensional elevation model of the tidal flat, and regularly calculate and record the elevation data of the observation points.

[0144] In this embodiment, an observation line along the slope direction is set in the three-dimensional elevation model of the tidal flat; a plurality of observation points are evenly set on each of the observation lines; in the three-dimensional elevation model of the tidal flat, the three-dimensional coordinates closest to the coordinates of the observation point are obtained, and the elevation data of the observation point is obtained through the closest three-dimensional coordinates. It includes: calculating the elevation gradient between adjacent observation points on the same observation line; setting a gradient threshold; when the elevation gradient exceeds the gradient threshold, the interval between the observation points on this observation line is changed to a second interval, and the second interval is smaller than the first interval.

[0145] Specifically, a plurality of observation lines along the slope direction are set on the three-dimensional elevation model, and a plurality of observation points are evenly set on each observation line at a first interval; in the three-dimensional elevation model, the three-dimensional coordinates closest to the coordinates of the observation point are found, and the elevation data of the observation point is obtained through the closest three-dimensional coordinates.

[0146] Further, in the three-dimensional elevation model, after finding the three-dimensional coordinates closest to the coordinates of the observation point and obtaining the elevation data of the observation point through the closest three-dimensional coordinates, the following is further executed: calculating the elevation gradient between adjacent observation points on the same observation line, and if the elevation gradient exceeds the gradient threshold, the interval between the observation points on this observation line is changed to a second interval, and the second interval is smaller than the first interval.

[0147] Building a tidal flat elevation monitoring model based on tidal flat photovoltaic video images. A tidal flat elevation monitoring system based on tidal flat photovoltaic video images can effectively improve the monitoring efficiency and accuracy of tidal flat changes.

[0148] It should be noted that it should be understood that the division of each module of the above system is only a division of logical functions. In actual implementation, it can be fully or partially integrated into a physical entity, or physically separated. And these modules can all be implemented in the form of software called by a processing element; they can also all be implemented in the form of hardware; or some modules can be implemented in the form of software called by a processing element, and some modules can be implemented in the form of hardware. For example, the x module can be a separately established processing element, or can be integrated in a certain chip of the above system. In addition, it can also be stored in the memory of the above system in the form of program code, and called and executed by a certain processing element of the above system to perform the functions of the above x module. The implementation of other modules is similar. In addition, these modules can be fully or partially integrated together or independently implemented. The processing element here can be an integrated circuit with signal processing capabilities. In the implementation process, each step of the above method or each of the above modules can be completed by the integrated logic circuit in the hardware of the processor element or the instructions in the form of software.

[0149] The above-mentioned modules may be one or more integrated circuits configured to implement the above methods. For example: one or more Application Specific Integrated Circuits (ASICs), or, one or more digital signal processors (DSPs), or, one or more Field Programmable Gate Arrays (FPGAs), etc. Again, when a certain above-mentioned module is implemented in the form of a processing element scheduling program code, the processing element may be a general-purpose processor, such as a Central Processing Unit (CPU) or other processors that can call program code. Again, these modules may be integrated together and implemented in the form of a system-on-a-chip (SOC).

[0150] Please refer to Figure 9 , which shows a schematic diagram of the principle structure of the tidal flat elevation monitoring device based on tidal flat photovoltaic video images according to an embodiment of the present application. As Figure 9 shown, this embodiment provides a tidal flat elevation monitoring device based on tidal flat photovoltaic video images. The tidal flat elevation monitoring device based on tidal flat photovoltaic video images includes: a processor 91 and a memory 92; the memory 92 is used to store a computer program; the processor 91 is connected to the memory 92 and is used to execute the computer program stored in the memory 92, so that the tidal flat elevation monitoring device based on tidal flat photovoltaic video images executes each step of the tidal flat elevation monitoring method based on tidal flat photovoltaic video images as described above.

[0151] Preferably, the memory may include a Random Access Memory (RAM), and may also include a non-volatile memory, such as at least one disk memory.

[0152] The above-mentioned processor may be a general-purpose processor, including a Central Processing Unit (CPU for short), a Network Processor (NP for short), etc.; it may also be a Digital Signal Processor (DSP for short), an Application Specific Integrated Circuit (ASIC for short), a Field Programmable Gate Array (FPGA for short), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0153] In summary, the method and system for monitoring the elevation of the beach surface based on the beach photovoltaic video image provided by this application have the following beneficial effects:

[0154] The method for monitoring the elevation of the beach surface based on the beach photovoltaic video image provided by this application selects a suitable location in the already built offshore beach photovoltaic field area. By installing a camera on the photovoltaic support, the characteristics of its fixed position can be utilized. The photovoltaic support can provide a relatively high installation position, and the field of view of the camera can cover a relatively large area of the beach. Unattended automated data collection is carried out, reducing the frequency and intensity of manual surveys. During the ebb and flow of tides or the process of erosion and deposition, the change of image information in the video has time criticality. Through scene recognition and frame extraction of pixel point changes, the key states of these dynamic change nodes can be extracted. In this application, by calculating the depth information, elevation feature points or image parallax of the pixel points in the image, the plane image data can be converted into the elevation data of the actual beach terrain. Using the extracted elevation data, a three-dimensional model covering the beach terrain can be generated, including tidal creeks, sandbars, and slope features, providing an intuitive overall morphology reconstruction. At the same time, by setting observation lines and observation points (refined to key areas) in the three-dimensional elevation model, the elevation monitoring of key change areas of the beach becomes more systematic and accurate. By regularly calculating and recording the elevation values of the observation points, dynamic comparative analysis can be carried out on the processes of beach deposition, erosion, regional expansion or retreat.

[0155] The above embodiments merely illustrate the principles and effects of this application, rather than limiting this application. Any person familiar with this technology can modify or change the above embodiments without departing from the spirit and scope of this application. Therefore, all equivalent modifications or changes completed by those with ordinary knowledge in the technical field without departing from the spirit and technical idea disclosed by this application should still be covered by the claims of this application.

Claims

1. A method for monitoring the elevation of a tidal flat based on tidal flat photovoltaic video images, characterized in that: The following steps are involved: Obtain tidal flat video of the target area to be monitored; Perform frame extraction processing based on the tidal flat video to obtain a tidal flat image key frame; Acquiring tidal flat surface elevation information according to the tidal flat image key frame; Performing three-dimensional modeling based on the tidal flat surface elevation information to generate a three-dimensional elevation model of the tidal flat; Observation lines and observation points are set in the three-dimensional elevation model of the tidal flat, and the elevation data of the observation points are calculated and recorded regularly.

2. The method for monitoring the elevation of tidal flats based on tidal flat photovoltaic video images according to claim 1 is characterized in that: Based on the tidal flat video, frame extraction processing is performed to obtain tidal flat image key frames including: Reading the tidal flat video frame by frame, preprocessing each frame of the tidal flat video image, and obtaining a first tidal flat video image; Constructing a scene classification prediction model based on the tidal flat video; Inputting the beach video image into the scene classification prediction model to obtain the scene type corresponding to the beach video image; If the current tidal flat video image is a tidal flat exposure scene, the frame corresponding to the tidal flat exposure scene is marked as a candidate key frame, and the proportion of changed pixels in each candidate key frame is calculated; If the proportion of changed pixels is greater than the change threshold, the frame is marked as a mudflat image key frame, otherwise the current frame is skipped.

3. The method for monitoring the elevation of tidal flats based on tidal flat photovoltaic video images according to claim 2 is characterized in that: Building a scene classification prediction model based on the tidal flat video includes: Extracting a number of tidal flat video image frames based on the tidal flat video; Marking is performed based on the tidal flat video image frame to obtain a scene label; the scene label includes: any one or more combinations of high tide period, low tide period, exposed tidal flat, and flooded tidal flat; A scene classification prediction model is constructed through a convolutional neural network, with image frames as input and scene labels as output, and the scene classification prediction model is trained.

4. The method for monitoring the elevation of tidal flats based on tidal flat photovoltaic video images according to claim 2 is characterized in that: The calculation of the ratio of changed pixels of each candidate key frame includes: Calculate the gray value difference between the corresponding pixels of the two adjacent frames; Obtain changed pixels according to filtering rules, and calculate the proportion of changed pixels; the filtering rules include: setting a gray value difference threshold; filtering changed pixels whose gray value difference is less than the gray value difference threshold; and counting other changed pixels when the gray value difference is greater than or equal to the gray value difference threshold. The calculation formula of the gray value difference is: difference=I t (x,y)-I t-1 (x,y) Among them, I t (x, y) represents the pixel value of the current frame; I t-1 (x,y) represents the pixel value of the previous frame.

5. The method for monitoring the elevation of tidal flats based on tidal flat photovoltaic video images according to claim 1 is characterized in that: Acquiring the tidal flat surface elevation information according to the tidal flat image key frame includes: Acquire a number of image features based on the tidal flat image key frame, and match the image features of the tidal flat image key frames captured by different probes in the same tidal flat area; Set a similarity threshold; Selecting feature points whose similarity is greater than the similarity threshold from the plurality of image features, and calculating the horizontal disparity of each feature point in the tidal flat image; generating a disparity map using the horizontal disparity; Acquire depth information of the disparity map by using a triangulation method, and convert the depth information into elevation information; The calculation formula of the horizontal disparity is: d = x1-x2 Wherein, d represents the horizontal disparity value; x1 and x2 represent the horizontal coordinates of the feature points in the two mudflat images respectively.

6. The method for monitoring the elevation of tidal flats based on tidal flat photovoltaic video images according to claim 5 is characterized in that: Generating a disparity map by using the horizontal disparity, acquiring depth information of the disparity map by using a triangulation method, and converting the depth information into elevation information comprises: Calculate the depth information of the disparity map based on the disparity information; the disparity information includes: disparity value, probe focal length, baseline distance between two probes, and probe depth value; The probe depth value is calibrated using the local elevation reference of the tidal flat to obtain the elevation information of the tidal flat surface; The calculation formula of the probe depth value is: Among them, Z represents the probe depth value; f represents the probe focal length; B represents the baseline distance between two probes; and d represents the parallax value.

7. The method for monitoring the elevation of tidal flats based on tidal flat photovoltaic video images according to claim 1 is characterized in that: The three-dimensional elevation model of the tidal flat is: Z=Z Among them, (X, Y, Z) represents the three-dimensional coordinates of the pixel point; Z represents the depth information; (u, v) represents the coordinates of the pixel point; (c x ,c y ) represents the coordinates of the principal point of the image calibrated by the camera.

8. The method for monitoring the elevation of tidal flats based on tidal flat photovoltaic video images according to claim 1 is characterized in that: Setting observation lines and observation points in the three-dimensional elevation model of the tidal flat, and regularly calculating and recording the elevation data of the observation points include: Setting an observation line along the slope direction in the three-dimensional elevation model of the tidal flat; A number of observation points are evenly arranged on each of the observation lines; In the tidal flat three-dimensional elevation model, the three-dimensional coordinates closest to the observation point coordinates are obtained, and the elevation data of the observation point is obtained through the closest three-dimensional coordinates.

9. The method for monitoring the elevation of tidal flats based on tidal flat photovoltaic video images according to claim 8 is characterized in that: In the tidal flat three-dimensional elevation model, obtaining the three-dimensional coordinates closest to the observation point coordinates, and obtaining the elevation data of the observation point through the closest three-dimensional coordinates includes: Calculating the elevation gradients of adjacent observation points on the same observation line; Set a gradient threshold; When the elevation gradient exceeds the gradient threshold, the interval of the observation points on the observation line is changed to a second interval, which is smaller than the first interval.

10. A tidal flat surface elevation monitoring system based on tidal flat photovoltaic video images, characterized in that: include: An acquisition module is used to acquire the tidal flat video of the target area to be monitored; A frame extraction module, used for performing frame extraction processing based on the beach video to obtain a beach image key frame; An elevation information acquisition module, used for acquiring the elevation information of the tidal flat surface according to the tidal flat image key frame; A model building module is used to perform three-dimensional modeling based on the elevation information of the tidal flat surface to generate a three-dimensional elevation model of the tidal flat; The elevation monitoring module is used to set observation lines and observation points in the three-dimensional elevation model of the tidal flat, and regularly calculate and record the elevation data of the observation points.

Images