Sea surface roughness inversion method based on image edge point density

Abstract

The invention discloses a sea surface roughness inversion method based on image edge point density, and the method comprises the steps: collecting a sea surface image under a specified shooting condition, and recording the real-time wind speed information; when images are collected, the average gray scale of the collected images is kept within an assumed interval [m, n]; image edge points are determined according to the threshold value th of the gradient module values, the sum of the gradient module values of all edge pixel points is divided by the number of the total pixel points of the image, the final calculation result serves as the image edge point density # imgabs0 #, and the sea surface roughness z0 at the corresponding wind speed at all the shooting moments is solved according to the wind speed at the high position of z meters; using data fitting to process image edge point density # imgabs1 # and sea surface roughness z0 data, and constructing a mapping relation between the image edge point density # imgabs2 # and the sea surface roughness z0 under the shooting condition; and calculating the image edge point density # imgabs3 # of the to-be-measured image, and obtaining a sea surface roughness inversion value z0'according to the constructed mapping relation between the image edge point density and the sea surface roughness.

Description

Technical Field

[0001] The invention relates to the field of digital image processing, and in particular to a sea surface roughness inversion method based on image edge point density. Background Art

[0002] The definition of sea surface roughness in aerodynamics is the height relative to sea level where the wind profile is zero. It is a physical quantity that characterizes the weakening effect of the roughness of the ocean surface on air flow. It is of great significance in marine engineering applications, such as port design and mooring stability conditions. Accurate estimation of sea surface roughness is of great help in inverting the wavelength and period of waves.

[0003] At present, visual observation technology has shown significant advantages in sea surface roughness estimation, and a series of studies have been carried out in the field of sea surface roughness visual observation technology at home and abroad, mainly focusing on inversion methods based on image features. Image features are parts of an image that are easily recognizable and different from other areas. Image features contain rich information. Among them, the area where the color, brightness or texture of the image changes significantly is called an edge. This image feature usually corresponds to the outline of an object. The image composed of edges is also called an edge image, and the pixels that make up these edges are called edge pixels. The sea surface roughness inversion method based on image features uses image features to reflect the wavelength, wave height and other parameters of the waves, and obtains the corresponding sea surface roughness by constructing a mapping model. However, there are still two problems: one is that the image feature calculation is complex; the other is that traditional algorithms usually ignore the intensity of edge image changes when calculating image features, and pay more attention to the frequency of edge images, which makes it difficult to accurately describe the state of the waves. Summary of the invention

[0004] In view of the problems caused by using the above-mentioned image features for sea surface roughness inversion, the present invention provides a sea surface roughness inversion method based on image edge point density. This method innovatively combines the number of edge points of the wave image with the edge image gradient amplitude, realizes the quantitative description of the wave characteristics, makes the mapping relationship of the sea surface roughness estimation model clearer, and further improves the interpretability and computational efficiency of the model.

[0005] The technical solution adopted by the present invention to solve its technical problem is:

[0006] A sea surface roughness inversion method based on image edge point density comprises the following steps:

[0007] Step 1), at a distance of b meters from the sea level, use a camera with a focal length of f and a resolution of i×j to collect images, where the angle between the camera optical axis and the sea level is a°. At the same time, record the real-time wind speed at a height of z meters.

[0008] Step 2), calculate the average grayscale i of the input sea surface grayscale image, and keep the average grayscale i within the assumed interval [m,n].

[0009] Step 3), calculate the gradient modulus of each pixel in the image. Define the pixel whose gradient modulus is greater than the threshold th as the edge pixel. Calculate the sum of the gradient modulus of all edge pixels and divide it by the total number of pixels in the image. The final calculation result is used as the image edge point density Used to reflect the wave state corresponding to the wave image.

[0010] Step 4), according to the wind speed at a height of z meters, calculate the sea surface roughness z at the corresponding wind speed at all shooting times 0 . Make every shot image They all correspond to the roughness of the sea surface at the time of shooting.

[0011] Step 5), use data fitting to process the image edge point density and sea surface roughness z 0 Data, construct the image edge point density under this shooting condition and sea surface roughness z 0 The mapping relationship.

[0012] Step 6), input the grayscale image of the sea wave that meets the acquisition requirements, repeat steps 2 to 3, calculate the image edge point density of the input image, and obtain the sea surface roughness inversion value z according to the constructed mapping relationship between the image edge point density and the sea surface roughness. 0 '. Finally, the sea surface roughness information is inverted using the sea surface images of the specified area.

[0013] Advantages or beneficial effects of the present invention:

[0014] (1) The present invention introduces the concept of image edge point density to solve the problem that the current traditional sea surface roughness visual observation technology cannot fully reflect the state of the waves due to ignoring the intensity of the edge pixel gradient. This method enhances the correlation between the input and output of the mapping relationship model and improves the accuracy of the algorithm.

[0015] (2) The image feature calculation method designed by the present invention is simple and the algorithm complexity is low, which is conducive to being loaded into a commonly used embedded platform, so that in subsequent engineering applications, it has the characteristics of low cost, high performance and high robustness. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 Flow chart of the method for inverting sea surface roughness based on image edge point density in an embodiment.

[0017] Figure 2 Schematic diagram of image acquisition.

[0018] Figure 3 To actually capture images of the sea surface.

[0019] Figure 4 is the image edge point density and the corresponding sea surface roughness (×10 -3 m) relationship diagram. DETAILED DESCRIPTION

[0020] The present invention is described in detail below in conjunction with the accompanying drawings and embodiments, examples of which are shown in the accompanying drawings. The embodiments described below with reference to the accompanying drawings are exemplary and intended to be used to explain the present invention, but should not be construed as limiting the present invention.

[0021] Example:

[0022] The sea surface roughness inversion method based on image edge point density proposed by the present invention is described below with reference to the accompanying drawings.

[0023] like Figure 1 As shown, a sea surface roughness inversion method based on image edge point density includes the following steps:

[0024] Step 1), such as Figure 2 As shown, the image acquisition device is set at 2.45 meters above the sea level, and uses a camera with a focal length of 35 mm and a resolution of 2448×2048 to collect images, wherein the angle between the camera optical axis and the sea level is 80°. At the same time, an anemometer is used to record the real-time wind speed at a height of about 2.45 meters above the sea level.

[0025] Step 2), input the collected sea surface grayscale image as follows Figure 3 As shown, the average grayscale i of the image is calculated and the average grayscale i is kept within the assumed interval [90,110].

[0026] Step 3), calculate the lateral gradient g of each pixel point respectively x , longitudinal gradient g y And the gradient modulus Define the pixel points with gradient modulus greater than 10 as edge pixels. Calculate the sum of the gradient modulus values ​​of all edge pixels and divide it by the total number of pixels in the image. The final calculation result is the image edge point density reflecting the wave state of the wave image.

[0027] Step 4) Calculate the average wind speed of the collected image corresponding to the 5-second period based on the real-time wind speed Based on the wind speed conversion relationship and wind speed data at different heights at sea The conversion formula of the average wind speed at sea at a height of 2.45 meters and 10 meters is as follows:

[0028]

[0029] in, is the average wind speed at a height of 10 meters. Combining the empirical formula of the average wind speed at a height of 10 meters and the roughness of the sea surface, the mapping formula of the average wind speed at a height of 2.45 meters and the roughness of the sea surface can be derived. Get the sea surface roughness z of the corresponding image in the image set 0 , the sea surface roughness result and the image edge point density One-to-one correspondence. The sea surface roughness mapping formula of wind speed at 2.45 meters and sea surface roughness is as follows:

[0030]

[0031] Step 5), using the polynomial fitting method, according to the sea surface roughness z 0 and the image edge point density The scatter plot of the two constructs a mapping relationship model, and the schematic diagram is as follows Figure 4 The corresponding mapping relationship under the image shooting conditions of this embodiment is as follows:

[0032] z 0 '=0.003614g 3 +0.000901g 2 -0.08585g+0.1718

[0033] in

[0034] Step 6), input the grayscale image of the sea wave that meets the acquisition requirements and repeat steps 2 to 3 to obtain the corresponding image edge point density. According to the constructed mapping relationship model between the image edge point density and the sea surface roughness, calculate the sea surface roughness inversion value z 0 '. Finally, the sea surface roughness information is inverted using the sea surface images of the specified area.

[0035] Compared with the common sea surface roughness image inversion method, the method of the present invention adds the gradient change of the wave image as the image feature, such as Figure 4 As shown in the figure, adding the gradient intensity information of the wave image strengthens the relationship between the image features and the sea surface roughness, and reduces the error of the mapping model between the image features and the sea surface roughness. This method also does not have the problem of taking too long to calculate the image features, and improves the speed and robustness of the sea surface roughness image inversion method.

[0036] The above-described embodiments are only preferred embodiments of the present invention and are not intended to limit the scope of implementation of the present invention. Therefore, all changes made according to the shape and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A sea surface roughness inversion method based on image edge point density, characterized in that: The steps include: Step 1), at a distance of b meters from the sea level, use a camera with a focal length of f and a resolution of i×j to collect images, where the angle between the camera optical axis and the sea level is a°. At the same time, at a height of z meters, use an anemometer to record the real-time wind speed. Step 2), calculate the average grayscale i of the input image and keep the average grayscale i in the assumed interval [m,n]. Step 3), calculate the gradient modulus of each pixel in the image. Calculate the sum of the gradient modulus of all edge pixels whose gradient modulus is greater than the threshold th, and divide it by the total number of pixels in the image. The final calculation result is used as the image edge point density Used to reflect the wave state corresponding to the wave image. Step 4), according to the wind speed at a height of z meters, solve the sea surface roughness z0 corresponding to the average wind speed at all shooting times. They all correspond to the roughness of the sea surface at the time of shooting. Step 5), use data fitting to process the image edge point density The edge point density of the image under this shooting condition is constructed by combining the sea surface roughness z0 data Mapping relationship with sea surface roughness z0. Step 6), input the grayscale image of the sea wave that meets the acquisition requirements, implement steps 2 to 3 on the input image, calculate the image edge point density of the input image, and obtain the sea surface roughness inversion value z0' based on the constructed mapping relationship between the image edge point density and the sea surface roughness. Finally, the sea surface roughness information is inverted using the sea surface image of the specified area.

2. The sea surface roughness inversion method based on image edge point density according to claim 1 is characterized in that In step 3), the definition of image edge point density is: the gradient modulus of the edge pixel point is regarded as the edge quality of the point, and the edge quality of the non-edge pixel point is 0. The image edge point density is the sum of the edge quality of the entire image divided by the total number of pixels, that is, the image edge point density.

3. The sea surface roughness inversion method based on image edge point density according to claim 1 is characterized in that In step 3), the gradient modulus of a pixel point in the image is the lateral gradient g of the point. x and longitudinal gradient g y The arithmetic square root of the sum of the squares of x and the longitudinal gradient g y The calculation formula is as follows: g x =[I(x+1,y)-I(x,y)+I(x,y)-I(x-1,y)] / 2 g y =[I(x,y+1)-I(x,y)+I(x,y)-I(x,y-1)] / 2 Where x is the column number of the point, y is the row number of the point, and I(x,y) represents the grayscale value of the pixel.

4. The sea surface roughness inversion method based on image edge point density according to claim 1 is characterized in that In step 5), the sea surface roughness z0 is correlated with the image edge point density The data are linked together to build a mapping relationship, and finally the sea surface roughness information can be inverted using the sea surface images of the specified area.

Images