CN114862726A - Tidal bore tide head line connection method based on gradient amplitude growth - Google Patents

Abstract

The invention discloses a tidal head line connection method based on gradient amplitude growth. Aiming at the feature of the large gradient amplitude of the tide line in the direction of the water flow when the tide arrives, the invention adopts the Sobel operator to calculate the gradient amplitude, and the operator can calculate the gradient amplitude of each pixel point in the direction of the water flow. By using the OTSU image segmentation method, the image is divided into foreground and background parts, and finally the pixel points are grown according to the magnitude of the gradient amplitude of the endpoint neighborhood of the fractured tide head line, so as to complete the tide head line connection. Finally, the contour with the largest area is the tidal head line of the tidal wave. This method can more accurately segment the tidal head line area of the tidal wave, and can accurately and effectively connect the tidal fracture parts, so as to improve the identification of the tidal head line. Accuracy to provide data support for tidal studies.

Description

A Tide Head Line Connection Method Based on Gradient Amplitude Growth

technical field

The invention belongs to the technical field of marine information, relates to a tidal surge observation method, and in particular relates to a tidal head line connection method based on gradient amplitude growth.

Background technique

Qiantang River tide is majestic and turbulent. It is a unique and spectacular natural landscape. Due to the different geographical environments of different river sections, different types of tide forms will be formed, such as the cross tide in Haining, the first-line tide in Yanguan, The returning tide of the Old Yancang and Meimei Dam attracts thousands of tourists every year to watch the tide. Some unnecessary casualties are caused. Therefore, exploring the law of tidal surge and studying its hydraulic characteristics and generation mechanism have important academic value and practical significance for tidal surge protection and protection.

Tide propagation speed is an important characteristic of tidal surge, which refers to the advancing speed of tidal head. The main influencing factors of tidal propagation speed are downstream tidal range, upstream water depth and runoff. The greater the water depth, the faster the tidal propagation. ; The larger the downstream tidal range, the faster the tidal propagation speed; the direction of runoff and tidal propagation is opposite, the larger the runoff, the slower the tidal propagation speed. The tidal propagation speed is closely related to tidal forecasting and shipping safety, and It is an important parameter for analyzing the hydraulic characteristics of tidal surge. The characteristics of tidal head height, tidal current velocity, and local tidal shape are all related to the tidal propagation velocity.

Patent No. CN202110430752.1 discloses a UAV-based machine vision surge velocity measurement method. The implementation of the method sequentially includes video image acquisition, grayscale and filtering processing, edge detection, expansion operation, and tide line extraction. There are six links in total and tidal velocity calculation. The tidal propagation velocity calculation method provided by this patent takes advantage of the UAV's mobility and wide detection range, which is conducive to obtaining more comprehensive tidal velocity information, and has the ability to measure a large spatial scale. , Tideline identification is accurate, non-contact security and high advantages, however, the edge detection method used in the detection of the tide line in the patent and the expansion operation used in the edge connection are both insufficient. The edge detection adopts the Canny detection algorithm, which detects the ripples or interference points on the river surface that do not belong to the tide line in the image as edges, which affects the detection results; although the expansion operation solves the problem of discontinuous edges to a certain extent, there are Obvious shortcomings: on the one hand, the edge break is too large to successfully connect the edges; on the other hand, the pixel point puffing operation only expands the pixel points of the broken part of the tide line, that is, the original pixel value of 255 pixels. The adjacent pixels are also set to 255, so the expanded pixels do not necessarily belong to part of the tide line, but only the broken part of the tide line is expanded by the pixel points to achieve the effect of filling the break. The expansion of the point will definitely make the area of the surging tide line expand, and what you get is not the original shape of the tide line, but the expanded tide line.

In view of the above problems, this patent proposes a tidal head line connection method based on gradient amplitude growth. In view of the large gradient amplitude of the tide head line in the Y-axis direction of the image when the tide arrives, that is, the direction of the water flow, the method The tide head line connection method uses the Sobel operator to calculate the gradient amplitude. This operator can calculate the magnitude of the gradient amplitude of each pixel on the Y-axis of the image. If the gradient amplitude is larger, it may belong to the tide head. By using the OTSU image segmentation method, the image is divided into foreground and background parts, and finally the pixel points are grown according to the magnitude of the gradient amplitude of the endpoint neighborhood of the fractured tide head line. Connect the tide head line at the tidal head line. Since there are other pixels in the image with the same high gradient amplitude, which are detected as the foreground part, finally the part with the largest contour area is screened out by comparing the size of the contour area. Tide line.

SUMMARY OF THE INVENTION

The purpose of the present invention is to make up for the deficiencies of the existing unmanned aerial vehicle (UAV) method for identifying the tide head line and calculating the tide propagation speed. The Sobel operator is used to calculate the gradient magnitude. This operator can calculate the magnitude of the gradient magnitude of each pixel on the Y-axis. The larger the gradient magnitude, the more likely it is part of the tide line. Iteratively uses OTSU The image segmentation method divides the image into foreground and background parts, and finally grows according to the magnitude of the gradient amplitude of the endpoint neighborhood of the fractured tidal head line, so as to complete the connection of the tidal head line, and finally filters the connection with the largest area. The domain is the tidal head line of the tidal wave. This method can more accurately segment the tidal head line area of the tidal wave, and can accurately and effectively connect the tidal fault parts, so as to improve the identification accuracy of the tidal head line, which is the best way to identify the tidal head line of the tidal wave. Research provides data support.

A tidal edge connection method based on gradient amplitude growth, as shown in Figure 1, includes the following steps:

Step 1: Grayscale and median filter processing are performed on the collected tide surge images.

Step 2: Use the Sobel operator to calculate the image amplitude gradient and convert the image format back to the original uint8 format.

Step 3: Calculate the image segmentation threshold using the OTSU image segmentation method and use the threshold to segment the tide surge image into foreground and background parts.

Step 4: Find all the right endpoints that meet the conditions in the image, which are the pixels where the tide head line breaks.

Step 5: Traverse each right endpoint, and select the pixels with the largest amplitude gradient of the three pixels on the right among the pixels in the adjacent 8 directions of the right endpoint as the target point for the growth of the tide head line.

Step 6: Continue to filter the pixels in the right neighborhood of the obtained target point. If there are no foreground pixels in its right neighborhood, select the pixel with the largest amplitude gradient as the target point for the next tide line growth, and repeat the process. Growth steps, the number of growth steps can be manually set according to the image effect.

Step 7: According to the contour area size of each independent foreground part in the image, the contour with the largest area is selected as the tide head line.

The specific steps of the step 1 are as follows: according to the different sensitivity of human vision to color, the B, G, and R channels of each pixel are respectively assigned different weight coefficients to weight and solve the gray value, which are sequentially regarded as the gray value of the entire image. Value, median filter is to arrange the gray values of pixels in the neighborhood of a certain point in the image in ascending order according to the neighborhood template, and calculate the median value of these gray values arranged from small to large, and use it as The grayscale value of this pixel.

The specific steps of the step 2 are: the Sobel operator is mainly used to obtain the first-order gradient of the digital image, the common application and physical meaning is edge detection, the pixel value of the image edge will change significantly, and the magnitude of the gradient amplitude represents The degree of change of the content in the image, first need to define a 3*3 convolution kernel, according to the characteristics of the obvious difference between the pixels before and after the tide head line when the tide arrives, select the convolution kernel as the Sobel operator in the Gy direction On the convolution template, the convolution kernel is convolved with the original image to obtain the vertical gradient value of each pixel point. The larger the value is, that is, the larger the gradient amplitude, the more likely it is the tide line of the tide. Since the result may be negative or greater than 255, it is also necessary to convert the data type to uint8 type to get a picture. A grayscale image with pixel values between 0 and 255.

The specific steps of the step 3 image segmentation are:

3-1: First calculate the histogram of the image, that is, all the pixels of the image are divided into 256 areas from 0 to 255, count the number of pixels falling in each area, and normalize the histogram

3-2: Define a threshold where i represents the classification, iterate from 0, count the proportion w0 of the pixels with gray levels of 0 to i in the entire image, and calculate the average gray level u0; count the gray levels of i to 255 The proportion w1 of the pixels of the whole image, and the average gray level u1 of the background pixels is counted;

3-3: Calculate the variance g=w0*w1*(u0-u1) of foreground pixels and background pixels;

3-4: i++; go to 3-2, end the iteration until i is 256, and get the i value corresponding to the maximum g;

3-5: Use the i value obtained in 3-4 as the threshold for the start of the iteration in 3-2, and repeat steps 3-2, 3-3, and 3-4 to obtain a new image segmentation threshold j;

3-6: Set the pixels whose gray level is greater than j to 255 in the image, and set the pixels below the threshold to 0, to obtain a picture with the foreground and background separated.

The specific steps of the step 4 are: traverse all the pixel points in the image, determine whether the three points on the right in the eight neighborhoods of the pixel point are all empty, if it is empty, it means that the point is the right endpoint, and use it as the right endpoint. Save, where the eight neighborhoods represent the adjacent pixels in eight directions connected by a pixel.

The specific steps of the step 5 are: traverse each right endpoint, and select the pixel point with the largest amplitude gradient of the three pixel points on the right in the eight neighborhoods of the right endpoint as the target point for the growth of the tide head line.

The specific steps of the step 6 are: continue to screen the three pixels on the right side of the target point obtained, if there are no foreground pixels in these three pixels, that is, the pixel value of the pixel is not 255, then select the maximum amplitude gradient. The pixel point of 1 is used as the target point of the next tide line growth, and the growth step is repeated, and the number of repetitions can be set manually.

The specific steps of the step 7 are: after image segmentation, the foreground part in the image includes the tide head line of the tide and other parts with high grayscale thresholds, and all contours in the image are found, and the contour is the one connected by the edges of the image. Overall, the area with the largest contour area is the tidal head line area.

The beneficial effects of the present invention are:

The present invention selects the convolution template in the Gy direction of the Sobel operator according to the characteristic that the vertical pixel points of the tidal surge change drastically. This method makes it easier to identify the tidal head line and avoids detecting other interfering pixel points to a greater extent.

By using the OTSU image segmentation method iteratively, the present invention obtains the tidal head line area of the tidal surge more accurately and improves the detection accuracy of the tidal head line.

The invention selects the pixel points with the largest amplitude gradient in the right neighborhood at the break of the tide head line as the growth target, which is more suitable for the tide head line, rather than a simple expansion operation, and expands the tide head line. area, and can complete the connection of the tidal head line when the fracture interval is large.

Description of drawings

Fig. 1 is the flow chart of the connection method of tidal head line based on gradient amplitude growth;

Fig. 2 is Sobel operator template diagram;

Figure 3 is the effect diagram of the connection and identification of the tide line;

Detailed ways

The present invention will be further illustrated below in conjunction with the accompanying drawings and specific embodiments.

Step 1: According to the different sensitivity of human vision to color, assign different weight coefficients to the B, G, and R channels of each pixel to calculate the gray value by weighting, and then use it as the gray value of the entire image. The calculation formula is as follows Formula (1) shows:

f(u,v)=0.1140*B(u,v)+0.5870*G(u,v)+0.2989*R(u,v) (1)

The median filter is to arrange the gray values of the pixels in the neighborhood of a certain point in the image in ascending order according to the neighborhood template, and calculate the median value of these gray values arranged from small to large, and use it as the pixel. The gray value of the point, its processing formula is shown in formula (2):

In the above formula, A is the neighborhood template window, median is the solution median, f(u, v) is the two-dimensional data sequence, and the final display effect is shown in Figure 3(a).

Step 2: The Sobel operator is mainly used to obtain the first-order gradient of a digital image. The common application and physical meaning is edge detection. The pixel value of the image edge will change significantly, and the magnitude of the gradient magnitude represents the degree of change in the content of the image. , First of all, a 3*3 convolution kernel needs to be defined. According to the obvious difference between the pixels before and after the tide head line when the tide arrives, the convolution kernel is selected as the convolution template of the Sobel operator in the Y-axis direction. , as shown in Figure 2(a), the Sobel operator shown in Figure 2(b) is suitable for situations where the image grayscale changes drastically under the X-axis. By convolving the convolution kernel with the original image, each pixel is obtained. The vertical gradient value of the point. The function of the convolution kernel is that when the difference between the upper and lower parts of the pixel is larger, the result is larger, that is, the larger the gradient amplitude, the more likely it is the tide line of the tide. , since the obtained result may be negative or greater than 255, it is also necessary to convert the data type to uint8 type, so that a grayscale image with a pixel value of 0 to 255 will be obtained, as shown in Figure 3(b).

Step 3: This process is to obtain the threshold value of image segmentation, and the specific steps are divided into the following steps:

3-1: First calculate the histogram of the image, that is, all the pixels of the image are divided into 256 areas from 0 to 255, count the number of pixels falling in each area, and normalize the histogram

3-2: Define a threshold where i represents the classification, iterate from 0, count the proportion w0 of the pixels with gray levels of 0 to i in the entire image, and calculate the average gray level u0; count the gray levels of i to 255 The proportion w1 of the pixels of the whole image, and the average gray level u1 of the background pixels is counted;

3-3: Calculate the variance g=w0*w1*(u0-u1) of foreground pixels and background pixels;

3-4: i++; go to 3-2, end the iteration until i is 256, and get the i value corresponding to the maximum g;

3-5: Use the i value obtained in 3-4 as the threshold for the start of the iteration in 3-2, and repeat steps 3-2, 3-3, and 3-4 to obtain a new image segmentation threshold;

3-6: Set the pixel points greater than the threshold in the image to 255, and set the pixels below the threshold to 0, to obtain a picture with the foreground and background separated. The result is shown in Figure 3(c).

Step 4: Traverse all the pixels in the image, and determine whether the three pixels on the right side of the eight adjacent pixels in the eight directions of the pixel are all backgrounds, that is, the pixel value is 0. If it is the background, it indicates the point. as the right endpoint and save it.

Step 5: Traverse each right endpoint, and select the pixel with the largest amplitude gradient of the three pixels on the right in the eight neighborhoods of the right endpoint as the target point for the growth of the tide head line.

Step 6: Continue to filter the pixels in the right neighborhood of the obtained target point. If there are no foreground pixels in its right neighborhood, select the pixel with the largest amplitude gradient as the target point for the next tide line growth, and repeat the process. For the growth step, the number of repetitions can be set manually. The larger the interval at the break of the tide head line, the more times the number of times can be set. This time, the number of times of growth is set to 20, and the result is shown in Figure 3(d).

Step 7: After image segmentation, the foreground part of the image includes the tide head line of the tide and other parts with high grayscale thresholds, find all the contours in the image, the contour is a whole connected by the edges of the image, and the contour area is the largest. The area is the tidal head line area, and the result is shown in Figure 3(e).

The above are the preferred embodiments of the present invention, but the protection scope of the present invention is not limited to this. Any person skilled in the art is within the technical scope disclosed by the present invention, and all changes or replacements are made according to the technical solutions of the present invention. should be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims (8)

1. a tidal head line connection method based on gradient amplitude growth, is characterized in that, comprises the following steps: Step 1: Grayscale and median filter processing are performed on the collected images; Step 2: Use the Sobel operator to calculate the image amplitude gradient and convert the image format back to the original uint8 form; Step 3: Iteratively use the OTSU image segmentation method to calculate the image segmentation threshold and use the threshold to segment the image into foreground and background parts; Step 4: Find all the right endpoints that meet the conditions in the image are the pixels at the break of the tide head line; Step 5: Traverse each right endpoint, and select the pixels with the largest amplitude gradient of the three pixels on the right among the pixels in the adjacent 8 directions of the right endpoint as the target point for the growth of the tide head line; Step 6: Continue to filter the pixels in the right neighborhood of the obtained target point. If there are no foreground pixels in its right neighborhood, select the pixel with the largest amplitude gradient as the target point for the next tide line growth, and repeat the process. Growth steps, the number of growth steps can be manually set according to the image effect; Step 7: According to the contour area size of each independent foreground part in the image, the contour with the largest area is selected as the tide head line. 2. a kind of tidal head line connection method based on gradient amplitude growth according to claim 1, is characterized in that the concrete step of described step 1 is: according to the different sensitivity of human vision to color, each The B, G, and R channels of the pixel are assigned different weight coefficients to solve the gray value by weighting, which are in turn used as the gray value of the entire image. The degree values are arranged in order from small to large, and the median value is calculated from the gray values arranged from small to large, which is used as the gray value of this pixel. 3. a kind of tidal head line connection method based on gradient amplitude growth according to claim 1, is characterized in that, the concrete step of described step 2 is: Sobel operator is mainly used to obtain the first step of digital image degree, select the convolution kernel as the convolution template of the Sobel operator in the Gy direction, and convolve the original image with the convolution kernel to obtain the vertical gradient amplitude of each pixel point. The function of the convolution kernel is when the The greater the difference between the upper and lower parts of the pixel, the greater the absolute value of the result, that is, the greater the gradient amplitude, the more likely it is the tide line of the tide. Since the result may be negative or greater than 255, so It is also necessary to convert the data type to uint8 type to obtain a grayscale image with pixel values between 0 and 255. 4. a kind of tidal head line connection method based on gradient amplitude growth according to claim 1, is characterized in that, the concrete step of described step 3 is: first need to calculate 0～255 gray through OTSU image segmentation method The first segmented image in the degree level is the gray level i of the foreground and background, and then calculates the gray level j of the foreground and background of the segmented image in the gray level from i to 255 by the OTSU image segmentation method, and the obtained j is Threshold for image segmentation, pixels whose gray level is greater than j are foreground pixels, otherwise they are background pixels. 5. a kind of tidal head line connection method based on gradient amplitude growth according to claim 1, is characterized in that, the concrete steps in described step 4 are: traverse all pixel points in the image, judge this pixel point Whether the three pixels on the right side of the adjacent pixels in the 8 directions are all backgrounds, that is, the pixel value is 0, if it is the background, it means that the point is the right endpoint, and it is saved. 6. a kind of tidal head line connection method based on gradient amplitude growth according to claim 1, is characterized in that, in described step 5, concrete steps are: traverse each right end point, select the eight neighbors of right end point The pixels with the largest amplitude gradient of the three pixels on the right side of the domain are used as the target points for the growth of the tide head line. 7. a kind of tidal head line connection method based on gradient amplitude growth according to claim 4, is characterized in that, described step 6 concrete steps are: continue to screen the pixel point in the right neighborhood of the target point obtained , if there is no foreground pixel in its right neighborhood, select the pixel with the largest amplitude gradient as the target point for the next tide line growth, repeat the growth step, the number of repetitions can be set artificially, and the tide line breaks The larger the interval, the more times you can set it. 8. The method for connecting tide head lines based on gradient amplitude growth according to claim 1, wherein the specific step of the step 7 is: through image segmentation, the foreground part in the image includes the tide Find all the contours in the image, the contour is a whole connected by the edges of the image, and the area with the largest contour area is the tide head line area.

Images