Coastal sediment particle size distribution measurement method and system based on image analysis

Abstract

The invention relates to the technical field of coast engineering and environment monitoring, in particular to a coast sediment particle size distribution measurement method and system based on image analysis, and the method comprises the steps: setting a calibration object in a target region, obtaining a surface sediment image of the target region, and carrying out the preprocessing of the surface sediment image, thereby obtaining a gray surface sediment image; performing two-dimensional wavelet transform on the gray surface sediment image to obtain a wavelet coefficient of each pixel point in the gray surface sediment image; calculating the median particle size of the sediment in the target area according to the wavelet coefficient of each pixel point; according to the calibration object and the surface sediment image, determining a conversion relation between a pixel size and an actual size of the surface sediment image, and establishing a corresponding relation between an image position coordinate and a geographic position coordinate; generating a two-dimensional distribution diagram of the sediment particle size according to the median particle size of the sediment based on the conversion relation and the corresponding relation; according to the invention, high-efficiency, high-precision and automatic measurement of coastal sediment particle size distribution can be realized, and sediment particle size distribution can be surveyed and mapped in real time in a large range.

Description

Technical Field

[0001] The present invention relates to the technical field of coastal engineering and environmental monitoring, and in particular to a method and system for measuring coastal sediment particle size distribution based on image analysis. Background Art

[0002] Sediment particle size analysis is a key technology in coastal engineering, environmental science, geology and other fields. Its core lies in revealing the characteristics of sediment particle size distribution, thereby providing a basis for understanding and analyzing the source, transport process and sedimentary environment of sediments. Sediment particle size refers to the size of sediment particles. Particle size distribution data can reflect information such as water body dynamic conditions, sedimentary environment and material source. Sediment particle size is usually classified into different grades such as gravel, sand, silt, clay, etc. Sedimentary environment and particle size distribution data are important parameters for analyzing the evolution of river, lake and marine environment.

[0003] At present, traditional methods for analyzing sediment particle size include screening analysis, sedimentation method and laser diffraction method. The screening method has a specific range for particle shape and particle size, is ineffective for silt and clay, and is time-consuming and labor-intensive to operate manually, with large single-measurement errors; the sedimentation method requires strict control of experimental conditions, and a single measurement cycle of >2 hours, which cannot be monitored in real time; the laser diffraction method relies on complex parameter settings and can provide high-precision results, but the equipment cost is expensive. Although existing image analysis methods have been used for sediment particle size analysis, they are still limited to microscopic images and are difficult to characterize macroscopic sediment characteristics. Using drones to quickly measure coastal topography and water depth has become a trend, but there is currently a lack of large-scale, efficient and accurate field sediment particle size measurement methods based on image analysis.

[0004] Therefore, there is an urgent need for a coastal sediment particle size distribution measurement method and system based on image analysis that can be applicable to complex outdoor environments, realize efficient, high-precision, low-cost automated measurement of coastal sediment particle size distribution, and dynamically map sediment particle size distribution over a large range (>1km). Summary of the invention

[0005] In order to solve the above technical problems, the present invention provides a method and system for measuring coastal sediment particle size distribution based on image analysis. Through the integration of unmanned aerial vehicle aerial photography and image analysis technology, it can realize efficient, high-precision and automated measurement of coastal sediment particle size distribution, and can dynamically and real-time map the sediment particle size distribution over a large range.

[0006] The present invention provides a method for measuring the particle size distribution of coastal sediment based on image analysis, comprising the following steps: S1. Set a calibration object in the target area, and use drone aerial photography to continuously take multi-angle photos to obtain the surface sediment image of the target area, and perform preprocessing to obtain a grayscale surface sediment image; S2, performing a two-dimensional wavelet transform on the grayscale surface sediment image to obtain the wavelet coefficients of each pixel in the grayscale surface sediment image at different scales; The calculation formula is as follows: ; Among them, W(x 0 ,y 0 ) represents the wavelet coefficient of the target pixel to be analyzed, (x 0 ,y 0 ) represents the coordinates of the target pixel to be analyzed, (x, y) represents the coordinates of any pixel in the grayscale surface sediment image, represents the mother wavelet function, represents the gray value of any pixel in the gray surface sediment image, and a represents the scaling parameter of the wavelet function; S3, calculating the median particle size of the sediment corresponding to each pixel point according to the weighted wavelet coefficients of different scales of each pixel point; S4, determining a conversion relationship between the pixel size of the surface sediment image and the actual size according to the actual size of the calibration object and the pixel size of the calibration object in the surface sediment image; S5. Establishing a corresponding relationship between the image position coordinates and the geographic location coordinates according to the GPS coordinates of the calibration object and the coordinates of the calibration object in the surface sediment image; S6, based on the conversion relationship between the pixel size and the actual size of the surface sediment image and the corresponding relationship between the image position coordinates and the geographical location coordinates, the real sediment particle size is generated according to the median particle size of the sediment corresponding to each pixel point; S7, reconstructing a three-dimensional model of the terrain of the target area through the SfM algorithm according to the surface sediment image, and obtaining three-dimensional point cloud data of the target area; S8. Convert the three-dimensional point cloud data of the target area into a digital elevation model of the terrain, extract the elevation information of the terrain and combine it with the actual sediment particle size to generate a two-dimensional distribution map of the sediment particle size.

[0007] Further, in S1, the surface sediment image is preprocessed to obtain a grayscale surface sediment image, wherein the preprocessing includes: distortion correction processing, grayscale processing, median filtering processing, illumination correction processing and peak detection processing; Among them, the Brown distortion model with adaptive parameter optimization is used for distortion correction; Light correction processing is performed through a dual-modal light correction strategy.

[0008] Furthermore, the mother wavelet function adopts the Mexican hat wavelet, and the calculation formula of the mother wavelet function is as follows: .

[0009] Furthermore, in S3, the median particle size of the sediment corresponding to each pixel is calculated based on the weighted wavelet coefficients of different scales of each pixel. The calculation formula is as follows: ; Among them, W n represents the nth wavelet coefficient, a n represents the nth scaling parameter, a n The value range is [2 0 ,2 7 ], D represents the median particle size of the sediment corresponding to each pixel, and n represents the total number of scales of the wavelet coefficients.

[0010] The present invention also provides a coastal sediment particle size distribution measurement system based on image analysis, which is used to implement the above-mentioned coastal sediment particle size distribution measurement method based on image analysis. The system includes the following modules: An image acquisition module is used to set calibration objects in the target area and obtain surface sediment images of the target area through continuous multi-angle photography by drones; An image processing module, connected to the image acquisition module, is used to pre-process the surface sediment image to obtain a grayscale surface sediment image; A wavelet transform module is connected to the image processing module and is used to perform a two-dimensional wavelet transform on the grayscale surface sediment image to obtain the wavelet coefficients of each pixel in the grayscale surface sediment image at different scales; The calculation formula is as follows: ; Among them, W(x 0 ,y 0 ) represents the wavelet coefficient of the target pixel to be analyzed, (x 0 ,y 0 ) represents the coordinates of the target pixel to be analyzed, (x, y) represents the coordinates of any pixel in the grayscale surface sediment image, represents the mother wavelet function, represents the gray value of any pixel in the gray surface sediment image, and a represents the scaling parameter of the wavelet function; The particle size calculation module is connected to the wavelet transform module and is used to calculate the median particle size of the sediment corresponding to each pixel point according to the weighted wavelet coefficients of different scales of each pixel point; The particle size conversion module is connected to the particle size calculation module, and is used to determine the conversion relationship between the pixel size and the actual size of the surface sediment image according to the actual size of the calibration object and the pixel size of the calibration object in the surface sediment image; and to establish the corresponding relationship between the image position coordinates and the geographical location coordinates according to the GPS coordinates of the calibration object and the coordinates of the calibration object in the surface sediment image; and based on the conversion relationship between the pixel size and the actual size of the surface sediment image and the corresponding relationship between the image position coordinates and the geographical location coordinates, generate the real sediment particle size according to the median particle size of the sediment corresponding to each pixel point; The visualization module is connected to the particle size conversion module and is used to reconstruct the three-dimensional model of the target area terrain through the SfM algorithm based on the surface sediment image to obtain the three-dimensional point cloud data of the target area; and convert the three-dimensional point cloud data of the target area into a digital elevation model of the terrain, and extract the elevation information of the terrain and combine it with the actual sediment particle size to generate a two-dimensional distribution map of the sediment particle size.

[0011] Furthermore, the image acquisition module includes a drone module: Drone modules include: GPS positioning unit, used to collect the longitude and latitude information of each image taken by the drone; Altitude sensor, used to control the drone to fly at a preset altitude in the target area.

[0012] Furthermore, the image processing module includes: A distortion correction unit, used for performing distortion correction on the captured image; A filtering processing unit, used for denoising the captured image; An illumination correction unit, used for performing illumination correction on the captured image; A grayscale processing unit, used for converting the corrected image into a grayscale image; The peak detection unit is used to perform peak detection on the grayscale image to determine the center position of the sediment particles.

[0013] The embodiments of the present invention have the following technical effects: The present invention combines the GPS position information of the UAV with the particle size information in the image by combining wavelet transform and image analysis technology to achieve high-precision mapping of sediment particle size over a large range. In the particle size mapping process, the sediment particle size data and the terrain elevation information are strictly aligned in space by mapping the GPS coordinates of the calibration object with the image coordinates, thus avoiding the registration error in traditional multi-sensor data fusion. The three-dimensional terrain data created by the SEM (Structure from Motion) method can be accurately converted to actual size, and the particle size distribution information of sediment particles can be obtained quickly and accurately. Compared with traditional methods such as screening and sedimentation, the measurement efficiency and accuracy are greatly improved, and the influence of terrain changes on particle size distribution is also reflected. Image analysis technology realizes non-contact measurement of sediment particles, avoiding particle damage or contamination that may be caused by contact in traditional methods, thereby ensuring the reliability of measurement results. The device has a high degree of automation, combined with advanced image acquisition and processing technology, the entire measurement process is highly automated, reducing manual intervention and human errors. Through image analysis technology, the sediment particle size distribution results can be displayed in the form of intuitive images or charts, which is convenient for researchers to conduct further data analysis and mining. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0015] Figure 1 is a flow chart of a method for measuring coastal sediment particle size distribution based on image analysis provided by an embodiment of the present invention; Figure 2 is a schematic diagram of a coastal surface sediment image taken by a drone provided by an embodiment of the present invention; Figure 3 is a schematic diagram of a mother wavelet function of a two-dimensional wavelet transform provided by an embodiment of the present invention; Figure 4 is a pixel map of sediment particle size generated by an embodiment of the present invention; Figure 5 is a two-dimensional distribution map of sediment particle size generated by an embodiment of the present invention; Figure 6 It is a structural schematic diagram of a coastal sediment particle size distribution measurement system based on image analysis provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0016] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be described clearly and completely below. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work belong to the scope of protection of the present invention.

[0017] The embodiment of the present invention provides a method for measuring the particle size distribution of coastal sediment based on image analysis. Figure 1 is a flow chart of a method for measuring the particle size distribution of coastal sediment based on image analysis provided by an embodiment of the present invention, see Figure 1 , specifically including: S1. Set a calibration object in the target area, and use drone aerial photography to continuously shoot at multiple angles to obtain the surface sediment image of the target area, and perform preprocessing to obtain a grayscale surface sediment image.

[0018] In some embodiments, before the drone takes photos, calibration objects are set on the beach. For example, calibration objects can be placed on the beach at intervals of 1 m, and the diameter of the calibration objects is 19 cm.

[0019] In some embodiments, Figure 2 is a schematic diagram of a coastal surface sediment image taken by a drone according to an embodiment of the present invention, see Figure 2 , the drone can perform high-resolution image acquisition tasks and is equipped with a GPS positioning system to ensure that each image taken has accurate latitude and longitude information. The drone moves along the coastline at a preset constant altitude and shoots images of the sediment surface vertically downward; the preset altitude can be 10m, and the resolution of each image is 3000×4000 pixels to ensure the stability and accuracy of the captured image. With the built-in GPS and altitude sensor, the drone flies on a preset trajectory and automatically collects images to ensure that the image covers the entire target area.

[0020] In some embodiments, the preprocessing includes: distortion correction processing, grayscale processing, median filtering processing, illumination correction processing and peak detection processing; Among them, the Brown distortion model with adaptive parameter optimization is used for distortion correction; Light correction processing is performed through a dual-modal light correction strategy (dynamic combination of histogram equalization and gamma correction).

[0021] Among them, since the images taken by drones may be distorted due to lens distortion, especially in the edge areas of the image, the distortion will distort the shape and size of the sediment particles, thus affecting the accuracy of the particle size estimation. The traditional Brown distortion model uses a fixed parameter (k 1=0.1), the present invention proposes a distortion gradient sensing mechanism, which dynamically adjusts the Brown model parameters (k 1 ∈[0.05,0.2]) to correct the image distortion. The model can effectively correct the radial distortion and tangential distortion of the lens, so that the sediment particles in the image can restore their true shape. Through distortion correction, the edges of the sediment particles in the image are clearer, providing a high-quality image foundation for subsequent particle size estimation.

[0022] Natural disturbances (such as driftwood and garbage) and uneven lighting can produce noise in the image, which will affect the accuracy of particle size estimation. The image is processed with 8-nearest neighbor median filtering. In a 3×3 window, if the central pixel is the edge point of the sediment particle, only the median calculation is performed on the non-edge pixel to avoid edge blurring caused by filtering. The outline of the sediment particles is clearer, and the accuracy of particle size estimation is significantly improved. Uneven lighting can cause some areas in the image to be too bright or too dark, affecting the accuracy of particle size estimation. By analyzing the bimodal characteristics of the sediment image histogram, a grayscale-gamma value lookup table (γ∈[0.45,0.75]) is established. For low-contrast areas (grayscale difference <30), histogram equalization is preferred to stretch the dynamic range; for dark details (grayscale value <50), gamma correction is used to retain the shadow characteristics of sediment particles.

[0023] Furthermore, since the Brown model corrects the image distortion, the stretching of the originally distorted area at the edge of the image may cause local uneven brightness. Histogram equalization can unify the global contrast, and gamma correction can repair the edge vignetting, thereby enhancing detail visibility and improving image quality.

[0024] After correction and denoising, the image is grayed. Since color images contain a lot of redundant information and color information is not needed in particle size estimation, graying can reduce the amount of image data and improve the efficiency of image processing. Graying introduces color space adaptive conversion. In view of the characteristics of sediment particles in sediment images being yellow-brown (RGB: 150~200, 100~150, 50~100), the weighted coefficient is dynamically adjusted (Y=0.299R+0.587G+0.114B) to calculate the gray value of each pixel. The grayed image is more suitable for subsequent processing such as wavelet transform, while reducing the amount of calculation and improving the processing speed.

[0025] After completing the correction, noise processing and grayscale processing, the center position of the sediment particles is further determined by peak detection: in the grayscale image, the center position of the sediment particles usually corresponds to the peak value of the grayscale value. The center position of the sediment particles is determined by calculating the wavelet coefficient of each pixel and combining it with the weighted average method. Peak detection can accurately identify the center of the sediment particles and provide an accurate reference point for particle size estimation.

[0026] S2. Perform two-dimensional wavelet transform on the grayscale surface sediment image to obtain the wavelet coefficients of each pixel in the grayscale surface sediment image at different scales.

[0027] Specifically, two-dimensional wavelet transform is a frequency analysis method that amplifies, reduces and translates the mother wavelet function and multiplies it with the signal to obtain the local characteristic quantity of the signal relative to a certain frequency domain at a certain moment, which is called the wavelet coefficient.

[0028] In some embodiments, a two-dimensional Mexican Hat wavelet is used for wavelet transform. Figure 3 is a schematic diagram of a mother wavelet function of a two-dimensional wavelet transform provided by an embodiment of the present invention, see Figure 3 , the calculation formula of the mother wavelet function is as follows: ; Perform two-dimensional wavelet transform on the grayscale surface sediment image to obtain the wavelet coefficient of each pixel in the grayscale surface sediment image. The calculation formula is as follows: ; Among them, W(x 0 ,y 0 ) represents the wavelet coefficient of the target pixel to be analyzed, (x 0 ,y 0 ) represents the coordinates of the target pixel to be analyzed, (x, y) represents the coordinates of any pixel in the grayscale surface sediment image, represents the mother wavelet function, f(x, y) represents the gray value of any pixel in the gray surface sediment image, and a represents the scaling parameter of the wavelet function.

[0029] S3. Calculate the median particle size of the sediment corresponding to each pixel by weighting the wavelet coefficients of different scales of each pixel.

[0030] In some embodiments, Figure 4 is a pixel map of sediment particle size generated by an embodiment of the present invention, see Figure 4 , the median particle size of the sediment corresponding to each pixel is calculated by weighting the wavelet coefficients of different scales at each pixel. The calculation formula is as follows: ; Among them, Wn represents the corresponding nth wavelet coefficient, a n represents the nth scaling parameter, D represents the median particle size of the sediment corresponding to each pixel, and n represents the total number of scales of the wavelet coefficient. In order to capture particles of different sizes, wavelet transform is performed at multiple scales, so the scaling parameter a n From 2 0 To 2 7 The scaling parameters change gradually, that is, the scaling parameters at small scales capture small particles, and the scaling parameters at large scales capture large particles.

[0031] S4. Determine a conversion relationship between the pixel size of the surface sediment image and the actual size according to the actual size of the calibration object and the pixel size of the calibration object in the surface sediment image.

[0032] In some embodiments, the size of the calibration object in the image is identified and measured, compared with the actual known size of the calibration object (e.g., 19 cm), the pixel size ratio in the image, that is, the actual length represented by each pixel, is calculated, and a conversion relationship between the measured size value in the image and the actual size is established.

[0033] S5. Establish a corresponding relationship between the image position coordinates and the geographic location coordinates according to the GPS coordinates of the calibration object and the coordinates of the calibration object in the surface sediment image.

[0034] In some embodiments, the GPS positioning coordinates of the calibration object are used to find its corresponding position in the image. By matching the GPS coordinates of the calibration object with its coordinates in the image, a conversion relationship between the image coordinate system and the geographic coordinate system is established. Based on this, any point in the image can be converted into corresponding geographic coordinates according to its position relative to the calibration object.

[0035] S6. Based on the conversion relationship between the pixel size and the actual size of the surface sediment image and the corresponding relationship between the image position coordinates and the geographical location coordinates, the real sediment particle size is generated according to the median particle size of the sediment corresponding to each pixel point.

[0036] S7. Reconstruct a three-dimensional model of the terrain of the target area according to the surface sediment image by using the SfM algorithm to obtain three-dimensional point cloud data of the target area.

[0037] In some embodiments, under complex terrain conditions (such as sand and gravel mixed beaches), it is difficult to accurately reflect terrain changes by relying solely on two-dimensional images, which will affect the accuracy of the particle size distribution map. Therefore, this solution uses drones to take multiple images at different heights to ensure that there is enough overlap between the images. The three-dimensional model of the terrain is reconstructed from these multi-viewpoint images through the Structure from Motion (SfM) algorithm. The SfM algorithm uses feature point matching and motion recovery structure technology in the image to generate high-precision three-dimensional point cloud data. The three-dimensional terrain model generated by the SfM technology can accurately reflect the changes in the terrain and provide a high-precision terrain background for the particle size distribution map.

[0038] S8. Convert the three-dimensional point cloud data of the target area into a digital elevation model of the terrain, extract the elevation information of the terrain and combine it with the actual sediment particle size to generate a two-dimensional distribution map of the sediment particle size.

[0039] In some embodiments, the terrain information generated by the SfM technology is fused with the particle size estimation results to generate a high-precision particle size distribution map. Under complex terrain conditions, the particle size distribution is not only affected by the characteristics of the particles themselves, but also closely related to terrain changes. Therefore, the fusion of terrain information and particle size information can more comprehensively reflect the actual situation of particle size distribution. The terrain elevation information generated by the SfM technology is combined with the particle size estimation results to generate a two-dimensional map of particle size distribution. In the two-dimensional map, not only the size of the particle size is displayed, but also the elevation changes of the terrain are marked, so that the particle size distribution map can intuitively reflect the impact of the terrain on the particle size distribution. Through the fusion of terrain information, the particle size distribution map not only shows the spatial distribution of particle size, but also reflects the impact of terrain changes on particle size distribution, making the particle size distribution map more accurate and intuitive. Elevation information (such as slope, curvature) can be directly associated with sediment particle size distribution, revealing the control mechanism of terrain on sediment transport and deposition (such as the enrichment of coarse particles on steep slopes).

[0040] In addition, the elevation information extracted by the digital elevation model (DEM) can be used to perform terrain correction on the two-dimensional distribution map (such as eliminating projection distortion), which solves the problem of traditional planar distribution maps ignoring terrain undulations and causing spatial expression distortion, and further improves the accuracy of sediment particle size distribution.

[0041] In some embodiments, Figure 5 is a two-dimensional distribution diagram of sediment particle size generated by an embodiment of the present invention, see Figure 5 , combined with the previously established size conversion relationship and geographic coordinate relationship, the median particle size of the sediment in each target area is converted to the actual size and mapped to the actual geographic location. Finally, based on the particle size values ​​of all target areas and their corresponding geographic coordinates, a two-dimensional distribution map is generated that can accurately show the changes in sediment particle size in different coastal areas.

[0042] Finally, in order to verify the accuracy of the system, this embodiment compares the actual on-site measurement data with the system calculation results. The experiment shows that the average particle size error calculated by this scheme is within 4.7%, which can meet the actual application requirements. The accuracy error calculation formula used in this embodiment is: ; Among them, E is the error percentage, D 测量 is the actual measured value, D 计算 Calculates values ​​for the system.

[0043] The present invention combines the GPS position information of the UAV with the particle size information in the image by combining wavelet transform and image analysis technology to achieve high-precision mapping of sediment particle size over a large range. In the particle size mapping process, the sediment particle size data and the terrain elevation information are strictly aligned in space by mapping the GPS coordinates of the calibration object with the image coordinates, thus avoiding the registration error in traditional multi-sensor data fusion. The three-dimensional terrain data created by the SEM (Structure from Motion) method can be accurately converted to actual size, and the particle size distribution information of sediment particles can be obtained quickly and accurately. Compared with traditional methods such as screening and sedimentation, the measurement efficiency and accuracy are greatly improved, and the influence of terrain changes on particle size distribution is also reflected. Image analysis technology realizes non-contact measurement of sediment particles, avoiding particle damage or contamination that may be caused by contact in traditional methods, thereby ensuring the reliability of measurement results. The device has a high degree of automation, combined with advanced image acquisition and processing technology, the entire measurement process is highly automated, reducing manual intervention and human errors. Through image analysis technology, the sediment particle size distribution results can be displayed in the form of intuitive images or charts, which is convenient for researchers to conduct further data analysis and mining.

[0044] The embodiment of the present invention provides a system for measuring the particle size distribution of coastal sediment based on image analysis. Figure 6 is a schematic diagram of a coastal sediment particle size distribution measurement system based on image analysis provided by an embodiment of the present invention, see Figure 6 The system is used to implement the above-mentioned coastal sediment particle size distribution measurement method based on image analysis. The system includes the following modules: An image acquisition module is used to obtain surface sediment images of the target area through continuous aerial photography by a drone; An image processing module, connected to the image acquisition module, is used to pre-process the surface sediment image to obtain a grayscale surface sediment image; A wavelet transform module is connected to the image processing module and is used to perform a two-dimensional wavelet transform on the grayscale surface sediment image to obtain the wavelet coefficients of each pixel in the grayscale surface sediment image at different scales; The particle size calculation module is connected to the wavelet transform module and is used to calculate the median particle size of the sediment corresponding to each pixel point according to the weighted wavelet coefficients of different scales of each pixel point; The particle size conversion module is connected to the particle size calculation module, and is used to determine the conversion relationship between the pixel size and the actual size of the surface sediment image according to the actual size of the calibration object and the pixel size of the calibration object in the surface sediment image; and to establish the corresponding relationship between the image position coordinates and the geographical location coordinates according to the GPS coordinates of the calibration object and the coordinates of the calibration object in the surface sediment image; and based on the conversion relationship between the pixel size and the actual size of the surface sediment image and the corresponding relationship between the image position coordinates and the geographical location coordinates, generate the real sediment particle size according to the median particle size of the sediment corresponding to each pixel point; The visualization module is connected to the particle size conversion module and is used to reconstruct a three-dimensional model of the target area terrain through the SfM algorithm according to the surface sediment image to obtain three-dimensional point cloud data of the target area; and convert the three-dimensional point cloud data of the target area into a digital elevation model of the terrain, and extract the elevation information of the terrain and generate a two-dimensional distribution map of the sediment particle size in combination with the actual sediment particle size.

[0045] In some embodiments, the image acquisition module includes a drone module: Drone modules include: GPS positioning unit, used to collect the longitude and latitude information of each image taken by the drone; Altitude sensor, used to control the drone to fly at a preset altitude in the target area.

[0046] In some embodiments, the image processing module includes: A distortion correction unit, used for performing distortion correction on the captured image; A filtering processing unit, used for denoising the captured image; An illumination correction unit, used for performing illumination correction on the captured image; A grayscale processing unit, used for converting the corrected image into a grayscale image; The peak detection unit is used to perform peak detection on the grayscale image to determine the center position of the sediment particles.

[0047] In some embodiments, the visualization module can separately plot the sediment particle size perpendicular to the coastline and along the coastline.

[0048] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the technical solutions of the embodiments of the present invention.

Claims

1. A method for measuring the particle size distribution of coastal sediment based on image analysis, characterized in that: The steps include: S1. Set a calibration object in the target area, and use drone aerial photography to continuously take multi-angle photos to obtain the surface sediment image of the target area, and perform preprocessing to obtain a grayscale surface sediment image; S2, performing a two-dimensional wavelet transform on the grayscale surface sediment image to obtain wavelet coefficients of each pixel in the grayscale surface sediment image at different scales; The calculation formula is as follows: ； Among them, W(x0,y0) represents the wavelet coefficient of the target pixel to be analyzed, (x0,y0) represents the coordinates of the target pixel to be analyzed, and (x,y) represents the coordinates of any pixel in the grayscale surface sediment image. represents the mother wavelet function, represents the gray value of any pixel in the gray surface sediment image, and a represents the scaling parameter of the wavelet function; S3, calculating the median particle size of the sediment corresponding to each pixel point according to the weighted wavelet coefficients of different scales of each pixel point; S4, determining a conversion relationship between the pixel size of the surface sediment image and the actual size according to the actual size of the calibration object and the pixel size of the calibration object in the surface sediment image; S5. Establishing a corresponding relationship between the image position coordinates and the geographic location coordinates according to the GPS coordinates of the calibration object and the coordinates of the calibration object in the surface sediment image; S6, based on the conversion relationship between the pixel size and the actual size of the surface sediment image and the corresponding relationship between the image position coordinates and the geographical location coordinates, the real sediment particle size is generated according to the median particle size of the sediment corresponding to each pixel point; S7, reconstructing a three-dimensional model of the terrain of the target area by using the SfM algorithm according to the surface sediment image to obtain three-dimensional point cloud data of the target area; S8. Convert the three-dimensional point cloud data of the target area into a digital elevation model of the terrain, extract the elevation information of the terrain and combine it with the actual sediment particle size to generate a two-dimensional distribution map of the sediment particle size.

2. The method for measuring the particle size distribution of coastal sediment based on image analysis according to claim 1, characterized in that: In S1, the surface sediment image is preprocessed to obtain a grayscale surface sediment image, wherein the preprocessing includes: distortion correction processing, grayscale processing, median filtering processing, illumination correction processing and peak detection processing; Among them, the Brown distortion model with adaptive parameter optimization is used for distortion correction; Light correction processing is performed through a dual-modal light correction strategy.

3. The method for measuring the particle size distribution of coastal sediment based on image analysis according to claim 1, characterized in that: The mother wavelet function adopts the Mexican hat wavelet, and the calculation formula of the mother wavelet function is as follows: 。 4. The method for measuring the particle size distribution of coastal sediment based on image analysis according to claim 3 is characterized in that: In S3, the median particle size of the sediment corresponding to each pixel is calculated by weighting the wavelet coefficients of different scales of each pixel, and the calculation formula is as follows: ； Among them, W n represents the nth wavelet coefficient, a n represents the nth scaling parameter, a n The value range is [2 0 ,2 7 ], D represents the median particle size of the sediment corresponding to each pixel, and n represents the total number of scales of the wavelet coefficients.

5. A coastal sediment particle size distribution measurement system based on image analysis, used to implement the coastal sediment particle size distribution measurement method based on image analysis as described in any one of claims 1 to 4, characterized in that: The system includes the following modules: An image acquisition module is used to set calibration objects in the target area and obtain surface sediment images of the target area through continuous multi-angle photography by drones; An image processing module, connected to the image acquisition module, for preprocessing the surface sediment image to obtain a grayscale surface sediment image; A wavelet transform module, connected to the image processing module, is used to perform a two-dimensional wavelet transform on the grayscale surface sediment image to obtain wavelet coefficients of each pixel in the grayscale surface sediment image at different scales; The calculation formula is as follows: ； Among them, W(x0,y0) represents the wavelet coefficient of the target pixel to be analyzed, (x0,y0) represents the coordinates of the target pixel to be analyzed, and (x,y) represents the coordinates of any pixel in the grayscale surface sediment image. represents the mother wavelet function, represents the gray value of any pixel in the gray surface sediment image, and a represents the scaling parameter of the wavelet function; A particle size calculation module, connected to the wavelet transform module, is used to calculate the median particle size of the sediment corresponding to each pixel point according to the weighted wavelet coefficients of different scales of each pixel point; A particle size conversion module is connected to the particle size calculation module, and is used to determine the conversion relationship between the pixel size and the actual size of the surface sediment image according to the actual size of the calibration object and the pixel size of the calibration object in the surface sediment image; and to establish a corresponding relationship between the image position coordinates and the geographical location coordinates according to the GPS coordinates of the calibration object and the coordinates of the calibration object in the surface sediment image; and based on the conversion relationship between the pixel size and the actual size of the surface sediment image and the corresponding relationship between the image position coordinates and the geographical location coordinates, generate the real sediment particle size according to the median particle size of the sediment corresponding to each pixel point; The visualization module is connected to the particle size conversion module and is used to reconstruct a three-dimensional model of the target area terrain through the SfM algorithm according to the surface sediment image to obtain three-dimensional point cloud data of the target area; and convert the three-dimensional point cloud data of the target area into a digital elevation model of the terrain, and extract the elevation information of the terrain and generate a two-dimensional distribution map of the sediment particle size in combination with the actual sediment particle size.

6. The coastal sediment particle size distribution measurement system based on image analysis according to claim 5 is characterized in that: The image acquisition module also includes a drone module: The drone module includes: GPS positioning unit, used to collect the longitude and latitude information of each image taken by the drone; Altitude sensor, used to control the drone to fly at a preset altitude in the target area.

7. The coastal sediment particle size distribution measurement system based on image analysis according to claim 5 is characterized in that: The image processing module also includes: A distortion correction unit, used for performing distortion correction on the captured image; A filtering processing unit, used for denoising the captured image; An illumination correction unit, used for performing illumination correction on the captured image; A grayscale processing unit, used for converting the corrected image into a grayscale image; The peak detection unit is used to perform peak detection on the grayscale image to determine the center position of the sediment particles.

Images