CN107833189A - The Underwater Target Detection image enchancing method of the limited self-adapting histogram equilibrium of contrast - Google Patents

Abstract

The invention discloses a contrast-limited adaptive histogram equalization underwater target detection image enhancement method. The steps include: calculating a 4-direction Sobel edge detector for a grayscale image corresponding to an original color image, a gradient image, and an adaptive Gain function; the original color image is non-linearly transformed from RGB space to HSI space; the brightness vector in the HSI space image is enhanced by a contrast-limited adaptive histogram equalization algorithm; the enhanced HSI space image is converted back to RGB space; the R, G, and B components in the enhanced RGB image are respectively subjected to generalized bounded multiplication based on the adaptive gain function to obtain an enhanced image based on the gradient information of the original image; the enhanced image is displayed; the enhanced image is Quantitative evaluation. The invention can make full use of the texture of the original image to realize image enhancement processing, so that the visual quality of the processed image is improved and the gradient information is rich.

Description

Image enhancement method for underwater target detection based on contrast-limited adaptive histogram equalization

technical field

The invention belongs to the field of image information processing, and in particular relates to an image enhancement method for underwater target detection with limited contrast and self-adaptive histogram equalization.

Background technique

Underwater target detection images have special conditions such as non-uniform brightness, low signal-to-noise ratio, and low contrast. Commonly used underwater target detection image enhancement algorithms are mainly divided into two categories: modifying the illumination of underwater images and suppressing image contrast to preserve image edges. , but it will inevitably reduce the visual quality of the detection image. The traditional image enhancement algorithm based on contrast enhancement has great limitations, such as histogram equalization to enhance the image globally, but increases the noise or introduces new noise. Local histogram equalization, that is, adaptive histogram equalization (AHE), although it overcomes the defect that global histogram equalization is difficult to adapt to local grayscale distribution, the artificial block effect is obvious after equalization. Therefore, contrast limited adaptive histogram equalization (CLAHE) has obvious advantages. Due to the influence of the optical characteristics of water and various particles, plankton and water flow in water, the research results of direct translation and transfer contrast-limited adaptive histogram equalization are still insufficient for underwater detection image enhancement. Combining the rich gradient information of the original image with the contrast-limited adaptive histogram equalization algorithm makes the details of the enhanced image richer and clearer, and the contrast and information entropy of the overall image can be effectively improved.

Contents of the invention

The technical problem to be solved by the present invention is to study a contrast-limited self-adaptive histogram in the face of the objective and practical requirements of accurate positioning and accurate description of underwater target detection images in environments with non-uniform brightness, low signal-to-noise ratio, and low contrast. The image equalization image enhancement method for underwater target detection realizes the denoising enhancement processing of the underwater target detection image, and improves the visual quality of the underwater target detection image.

The present invention adopts following scheme to realize:

The underwater object detection image enhancement method for contrasting restricted adaptive histogram equalization includes the following steps:

Step 1: Obtain the original color image of underwater target detection;

Step 2: Calculate the 4-direction Sobel edge detector, gradient image, and adaptive gain function of the grayscale image corresponding to the original color image;

Step 3: Transform the original color image from RGB space to HSI space through nonlinear transformation;

Step 4: applying a contrast-limited adaptive histogram equalization algorithm to the luminance vector in the HSI space image for enhancement processing;

Step 5: converting the enhanced HSI space image back to RGB space;

Step 6: Perform generalized bounded multiplication based on the adaptive gain function for the R, G, and B components of the enhanced RGB image, respectively, to obtain an enhanced image based on the gradient information of the original image;

Step 7: Transform the R, G, and B components of the enhanced image from the [0,1] range to the [0,255] range, and display the enhanced image;

Step 8: Quantitatively evaluate the enhanced image in terms of mean value, contrast, information entropy and color scale.

Further, said step two,

Calculate the grayscale image Gray(I) corresponding to the original color image I; use the characteristic that the human eye is sensitive to high-frequency information such as edges, and select the Sobel operator with noise robustness to obtain the edge gradient image; in the traditional Sobel operator filter , that is, on the basis of the 0 ^° and 90 ^° directions, add two diagonal directions, that is, filter in the 45° and 135° directions;

The Sobel edge detector mask in four directions is defined as:

Assuming that Z(i,j) is defined as the 3×3 image neighborhood of pixel point (i,j), then Z(i,j) can be expressed as:

Among them, z(i, j) is defined as the gray value of the pixel point (i, j);

The gradient vector of the 4 directions of the pixel point (i,j) can be defined as:

G _k (i,j)=∑∑z(i+m-1,j+n-1)×S _k (m,n),k=1,2,3,4

The gradient image of pixel point (i, j) can be defined as:

The gradient image is normalized to:

Among them, δ ₁ and δ ₂ are small perturbations to ensure g _n (i,j)∈(0,1);

The adaptive gain function λ(i,j) at the pixel point (i,j) is expressed as:

Among them, a and b are adjustable positive variables, which are used to make the adaptive gain function λ(i,j) mean

Further, the step three,

The original color image is nonlinearly transformed from RGB space to HSI space, and the conversion formula is:

in,

Before conversion, the values of R, G, and B are normalized to [0,1]; after conversion, S and I components are in the range of [0,1], and H is in the range of [0,360].

Further, the step four,

The luminance vector I in the HSI space image is enhanced by contrast-limited adaptive histogram equalization algorithm, and the color information (H, S) remains unchanged.

Further, the step five,

The enhanced HSI space image is converted back to RGB space, and the converted RGB space image is represented as R'G'B', and the conversion formula is:

(a) RG area (0°≤H<120°):

(b) GB area (120°≤H<240°):

H=H-120°

(c) BR zone (240°≤H<360°):

H=H-240°

Further, the step six,

Perform generalized bounded multiplication based on the adaptive gain function for the R', G', and B' components in the enhanced RGB image, and obtain the enhanced image R″G″B” based on the gradient information of the original image; generalized bounded multiplication The operation is expressed as:

Further, the step seven,

Transform the R, G, and B components of the enhanced image from the [0,1] range to the [0,255] range, and display the enhanced image;

The final output image can be expressed as:

in,

Further, the step eight,

Quantitatively evaluate the RGB _out of the enhanced image from the aspects of mean value, contrast, information entropy and color scale, and the related quantitative evaluation index function is expressed as:

mean: Wherein, μ _R , μ _G and μ _B are respectively the mean value of RGB _out three-channel color components;

Contrast: In the formula, P(i,j; d,θ _k ) is the gray level co-occurrence matrix; θ _k is the angle between pixels, θ _k =(k-1)×45°, k=1,2,3,4;

Information entropy:

Color scale: Among them, α=RG, β=(R+G)/2-B; μ _α and μ _β are the mean values of α and β respectively, and σ _α and σ _β are the standard deviations of α and β respectively.

Explanation on issues related to image enhancement of contrast-limited adaptive histogram equalization algorithm:

(1) Contrast There are two parameters in the limited adaptive histogram equalization algorithm: clipping coefficient (clip limit, CL) and image block size (block size, BZ). As the shear factor increases, the image brightness increases and the image smoothness increases; as the image block size increases, the image dynamic range increases. However, the image quality enhancement effect mainly depends on the shear factor rather than the image block size. In the actual application process, these two parameters need to be set reasonably.

(2) According to the needs of contrast enhancement, the parameters a and b in the adaptive gain function λ(i,j) can be adjusted appropriately in order to obtain enhanced images with different contrasts.

(3) When quantitatively evaluating the enhanced image, high contrast should not be required only, but contrast, information entropy and color information should be considered comprehensively.

The benefits achieved by the present invention are:

The method of the invention can only use the information of a single underwater target detection image with non-uniform brightness, low signal-to-noise ratio, and low contrast to perform contrast-limited adaptive histogram equalization denoising and enhancement processing on the image. First, the contrast-limited adaptive histogram equalization enhancement process is performed through the HSI space image, and then the rich gradient information of the image itself is used for adaptive gain. Enhanced images for testing. The 4-direction Sobel edge detector used in the present invention can make full use of the rich gradient information of the image itself to realize image enhancement processing, so that the visual quality of the processed image is improved and the texture information is rich.

Description of drawings

Fig. 1 is a control flow chart of the contrast-limited adaptive histogram equalization method for underwater target detection image enhancement in the present invention.

Detailed ways

The present invention will be further described below in conjunction with the accompanying drawings. The following examples are only used to illustrate the technical solution of the present invention more clearly, but not to limit the protection scope of the present invention.

Referring to Figure 1, the present invention is a method for enhancing underwater target detection images with contrast-limited adaptive histogram equalization. The overall flow chart is shown in Figure 1, and the specific implementation steps are as follows:

Step 1: Obtain the original color image I of the underwater target.

Step 2: Calculate the grayscale image Gray(I) corresponding to the original color image I. Taking advantage of the sensitivity of human eyes to high-frequency information such as edges, a Sobel operator with certain noise robustness is selected to obtain edge gradient images. On the basis of traditional Sobel operator filtering (0° and 90° directions), two diagonal directions (45° and 135° directions) are added to enhance the ability to smooth noise.

The Sobel edge detector mask in four directions is defined as:

Assuming that Z(i,j) is defined as the 3×3 image neighborhood of pixel point (i,j), then Z(i,j) can be expressed as:

Among them, z(i, j) is defined as the gray value of the pixel point (i, j).

The gradient vector of the 4 directions of the pixel point (i,j) can be defined as:

G _k (i,j)=∑∑z(i+m-1,j+n-1)×S _k (m,n),k=1,2,3,4

The gradient image of pixel point (i, j) can be defined as:

The gradient image is normalized to:

Among them, δ ₁ and δ ₂ are small perturbations to ensure g _n (i,j)∈(0,1).

In order to obtain an image with rich gradient information, the adaptive gain function λ(i,j) at the pixel point (i,j) can be expressed as:

Among them, a and b are adjustable positive variables to ensure the mean value of the adaptive gain function λ(i,j)

Step 3: Transform the original color image from RGB space to HSI space through nonlinear transformation, the conversion formula is:

in,

Before conversion, the values of R, G, B should be normalized to [0,1]. Then after conversion, S and I components are in the range of [0,1], and H is in the range of [0,360].

Step 4: Apply the contrast-limited adaptive histogram equalization algorithm to the luminance vector (I) in the HSI space image for enhancement processing, and the color information (H, S) remains unchanged.

Suppose the enhanced luminance vector is denoted as I', and the enhanced HSI spatial image is denoted as HSI'.

Step 5: Convert the enhanced HSI space image back to the RGB space, let the converted RGB space image be expressed as R'G'B', and the conversion formula is:

(a) RG area (0°≤H<120°):

(b) GB area (120°≤H<240°):

H=H-120°

(c) BR zone (240°≤H<360°):

H=H-240°

Step 6: Perform generalized bounded multiplication based on the adaptive gain function for the R', G', and B' components of the enhanced RGB image, respectively, to obtain the enhanced image R"G"B" based on the gradient information of the original image. The generalized bounded multiplication operation can be expressed as:

Step 7: Transform the R, G, and B components of the enhanced image from the [0,1] range to the [0,255] range, and display the enhanced image.

The final output image can be expressed as:

in,

Step 8: Quantitatively evaluate the RGB _out of the enhanced image in terms of mean value, contrast, information entropy, and color scale. The related quantitative evaluation index function is expressed as:

mean: Among them, μ _R , μ _G and μ _B are respectively the mean values of the RGB _out three-channel color components.

Contrast: In the formula, P(i,j; d,θ _k ) is the gray level co-occurrence matrix; θ _k is the angle between pixels, θ _k =(k-1)×45°, k=1,2,3,4.

Information entropy:

Color scale: Among them, α=RG, β=(R+G)/2-B; μ _α and μ _β are the mean values of α and β respectively, and σ _α and σ _β are the standard deviations of α and β respectively.

Explanation on issues related to image enhancement of contrast-limited adaptive histogram equalization algorithm:

(1) Contrast There are two parameters in the limited adaptive histogram equalization algorithm: clipping coefficient (clip limit, CL) and image block size (block size, BZ). As the shear factor increases, the image brightness increases and the image smoothness increases; as the image block size increases, the image dynamic range increases. However, the image quality enhancement effect mainly depends on the shear factor rather than the image block size. In the actual application process, these two parameters need to be set reasonably.

(2) According to the needs of contrast enhancement, the parameters a and b in the adaptive gain function λ(i,j) can be adjusted appropriately in order to obtain enhanced images with different contrasts.

(3) When quantitatively evaluating the enhanced image, high contrast should not be required only, but contrast, information entropy and color information should be considered comprehensively.

The above is only a preferred embodiment of the present invention, it should be pointed out that for those of ordinary skill in the art, without departing from the technical principle of the present invention, some improvements and modifications can also be made. It should also be regarded as the protection scope of the present invention.

Claims (8)

1. Contrast the underwater target detection image enhancement method of limited adaptive histogram equalization, comprising the following steps: Step 1: Obtain the original color image of underwater target detection; Step 2: Calculate the 4-direction Sobel edge detector, gradient image, and adaptive gain function of the grayscale image corresponding to the original color image; Step 3: Transform the original color image from RGB space to HSI space through nonlinear transformation; Step 4: applying a contrast-limited adaptive histogram equalization algorithm to the luminance vector in the HSI space image for enhancement processing; Step 5: converting the enhanced HSI space image back to RGB space; Step 6: Perform generalized bounded multiplication based on the adaptive gain function for the R, G, and B components of the enhanced RGB image, respectively, to obtain an enhanced image based on the gradient information of the original image; Step 7: Transform the R, G, and B components of the enhanced image from the [0,1] range to the [0,255] range, and display the enhanced image; Step 8: Quantitatively evaluate the enhanced image in terms of mean value, contrast, information entropy and color scale. 2. the underwater target detection image enhancement method of contrast limited adaptive histogram equalization according to claim 1, is characterized in that: described step 2, Calculate the grayscale image Gray(I) corresponding to the original color image I; use the characteristic that the human eye is sensitive to high-frequency information such as edges, and select the Sobel operator with noise robustness to obtain the edge gradient image; in the traditional Sobel operator filter , that is, on the basis of the 0° and 90° directions, add two diagonal directions, that is, filtering in the 45° and 135° directions; The Sobel edge detector mask in four directions is defined as: <mrow><msub><mi>S</mi><mn>1</mn></msub><mo>=</mo><mfenced open = "(" close = ")"><mtable><mtr><mtd><mrow><mo>-</mo><mn>1</mn></mrow></mtd><mtd><mn>0</mn></mtd><mtd><mn>1</mn></mtd></mtr><mtr><mtd><mrow><mo>-</mo><mn>2</mn></mrow></mtd><mtd><mn>0</mn></mtd><mtd><mn>2</mn></mtd></mtr><mtr><mtd><mrow><mo>-</mo><mn>1</mn></mrow></mtd><mtd><mn>0</mn></mtd><mtd><mn>1</mn></mtd></mtr></mtable></mfenced></mrow> <mrow><msub><mi>S</mi><mn>2</mn></msub><mo>=</mo><mfenced open = "(" close = ")"><mtable><mtr><mtd><mrow><mo>-</mo><mn>1</mn></mrow></mtd><mtd><mrow><mo>-</mo><mn>2</mn></mrow></mtd><mtd><mrow><mo>-</mo><mn>1</mn></mrow></mtd></mtr><mtr><mtd><mn>0</mn></mtd><mtd><mn>0</mn></mtd><mtd><mn>0</mn></mtd></mtr><mtr><mtd><mn>1</mn></mtd><mtd><mn>2</mn></mtd><mtd><mn>1</mn></mtd></mtr></mtable></mfenced></mrow> <mrow><msub><mi>S</mi><mn>3</mn></msub><mo>=</mo><mfenced open = "(" close = ")"><mtable><mtr><mtd><mn>0</mn></mtd><mtd><mn>1</mn></mtd><mtd><mn>2</mn></mtd></mtr><mtr><mtd><mrow><mo>-</mo><mn>1</mn></mrow></mtd><mtd><mn>0</mn></mtd><mtd><mn>1</mn></mtd></mtr><mtr><mtd><mrow><mo>-</mo><mn>2</mn></mrow></mtr>mtd><mtd><mrow><mo>-</mo><mn>1</mn></mrow></mtd><mtd><mn>0</mn></mtd></mtr></mtable></mfenced></mrow> <mrow><msub><mi>S</mi><mn>4</mn></msub><mo>=</mo><mfenced open = "(" close = ")"><mtable><mtr><mtd><mn>2</mn></mtd><mtd><mn>1</mn></mtd><mtd><mn>0</mn></mtd></mtr><mtr><mtd><mn>1</mn></mtd><mtd><mn>0</mn></mtd><mtd><mrow><mo>-</mo><mn>1</mn></mrow></mtd></mtr><mtr><mtd><mn>0</mn></mtd><mtd><mrow><mo>-</mo><mn>1</mn></mrow></mtd><mtd><mrow><mo>-</mo><mn>2</mn></mrow></mtd></mtr></mtable></mfenced></mrow> Assuming that Z(i,j) is defined as the 3×3 image neighborhood of pixel point (i,j), then Z(i,j) can be expressed as: <mrow><mi>Z</mi><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>)</mo></mrow><mo>=</mo><mfenced open = "(" close = ")"><mtable><mtr><mtd><mrow><mi>z</mi><mrow><mo>(</mo><mi>i</mi><mo>-</mo><mn>1</mn><mo>,</mo><mi>j</mi><mo>-</mo><mn>1</mn><mo>)</mo></mrow></mrow></mtd><mtd><mrow><mi>z</mi><mrow><mo>(</mo><mi>i</mi><mo>-</mo><mn>1</mn><mo>,</mo><mi>j</mi><mo>)</mo></mrow></mrow></mtd><mtd><mrow><mi>z</mi><mrow><mo>(</mo><mi>i</mi><mo>-</mo><mn>1</mn><mo>,</mo><mi>j</mi><mo>+</mo><mn>1</mi>mn><mo>)</mo></mrow></mrow></mtd></mtr><mtr><mtd><mrow><mi>z</mi><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>-</mo><mn>1</mn><mo>)</mo>mo></mrow></mrow></mtd><mtd><mrow><mi>z</mi><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>)</mo></mrow></mrow></mtd><mtd><mrow><mi>z</mi><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>+</mo><mn>1</mn><mo>)</mo></mrow></mrow></mtd></mtr><mtr><mtd><mrow><mi>z</mi><mrow><mo>(</mo><mi>i</mi><mo>+</mo><mn>1</mn><mo>,</mo><mi>j</mi><mo>-</mo><mn>1</mn><mo>)</mo></mrow></mrow></mtd><mtd><mrow><mi>z</mi><mrow><mo>(</mo><mi>i</mi><mo>+</mo><mn>1</mn><mo>,</mo><mi>j</mi><mo>)</mo></mrow></mrow></mtd><mtd><mrow><mi>z</mi><mrow><mo>(</mo><mi>i</mi><mo>+</mo><mn>1</mn><mo>,</mo><mi>j</mi><mo>+</mo><mn>1</mn><mo>)</mo></mrow></mrow></mtd></mtr></mtable></mfenced></mrow> Among them, z(i, j) is defined as the gray value of the pixel point (i, j); The gradient vector of the 4 directions of the pixel point (i,j) can be defined as: G _k (i,j)=∑∑z(i+m-1,j+n-1)×S _k (m,n),k=1,2,3,4 The gradient image of pixel point (i, j) can be defined as: <mrow><mi>g</mi><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>)</mo></mrow><mo>=</mo><msqrt><mrow><munderover><mo>&amp;Sigma;</mo><mrow><mi>k</mi><mo>=</mo><mn>1</mn></mrow><mn>4</mn></munderover><msubsup><mi>G</mi><mi>k</mi><mn>2</mn></msubsup><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>)</mo></mrow></mrow></msqrt></mrow> The gradient image is normalized to: <mrow><msub><mi>g</mi><mi>n</mi></msub><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>)</mo></mrow><mo>=</mo><mfrac><mrow><mi>log</mi><mrow><mo>(</mo><mi>g</mi><mo>(</mo><mrow><mi>i</mi><mo>,</mo><mi>j</mi></mrow><mo>)</mo><mo>+</mo><mn>1</mn><mo>+</mo><msub><mi>&amp;delta;</mi><mn>1</mn></msub><mo>)</mo></mrow></mrow><mrow><mi>log</mi><mrow><mo>(</mo><mi>max</mi><mo>(</mo><mrow><mi>g</mi><mrow><mo>(</mo><mrow><mi>i</mi><mo>,</mo><mi>j</mi></mrow><mo>)</mo></mrow></mrow><mo>)</mo><mo>+</mo><msub><mi>&amp;delta;</mi><mn>2</mn></msub><mo>)</mo></mrow></mrow></mfrac></mrow> Among them, δ ₁ and δ ₂ are small perturbations to ensure g _n (i,j)∈(0,1); The adaptive gain function λ(i,j) at the pixel point (i,j) is expressed as: <mrow><mi>&amp;lambda;</mi><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>)</mo></mrow><mo>=</mo><msup><mn>2</mn><mrow><mo>&amp;lsqb;</mo><mi>a</mi><mo>&amp;times;</mo><msub><mi>g</mi><mi>n</mi></msub><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>)</mo></mrow><mo>&amp;rsqb;</mo></mrow></msup><mo>+</mo><mi>b</mi></mrow> Among them, a and b are adjustable positive variables, which are used to make the adaptive gain function λ(i,j) mean 3. the underwater target detection image enhancement method of contrast limited adaptive histogram equalization according to claim 1, is characterized in that: described step 3, The original color image is nonlinearly transformed from RGB space to HSI space, and the conversion formula is: <mfenced open = "{" close = ""><mtable><mtr><mtd><mrow><mi>H</mi><mo>=</mo><mfenced open = "{" close = ""><mtable><mtr><mtd><mrow><mi>&amp;theta;</mi><mo>,</mo></mrow></mtd><mtd><mrow><mi>i</mi><mi>f</mi></mrow></mtd><mtd><mrow><mi>B</mi><mo>&amp;le;</mo><mi>G</mi></mrow></mtd></mtr><mtr><mtd><mrow><mn>360</mn><mo>-</mo><mi>&amp;theta;</mi>mi><mo>,</mo></mrow></mtd><mtd><mrow><mi>i</mi><mi>f</mi></mrow></mtd><mtd><mrow><mi>B</mi><mo&gt;&gt;</mo><mi>G</mi></mrow></mtd></mtr></mtable></mfenced></mrow></mtd></mtr><mtr><mtd><mrow><mi>S</mi><mo>=</mo><mn>1</mn><mo>-</mo><mfrac><mn>3</mn><mrow><mi>R</mi><mo>+</mo><mi>G</mi><mo>+</mo><mi>B</mi></mrow></mfrac><mo>&amp;lsqb;</mo><mi>min</mi><mrow><mo>(</mo><mi>R</mo>mi><mo>,</mo><mi>G</mi><mo>,</mo><mi>B</mi><mo>)</mo></mrow><mo>&amp;rsqb;</mo></mrow></mtd></mtr><mtr><mtd><mrow><mi>I</mi><mo>=</mo><mfrac><mn>1</mn><mn>3</mn></mfrac><mrow><mo>(</mo><mi>R</mi><mo>+</mo><mi>G</mi><mo>+</mo><mi>B</mi><mo>)</mo></mrow></mrow></mtd></mtr></mtable></mfenced> in, Before conversion, the values of R, G, and B are normalized to [0,1]; after conversion, S and I components are in the range of [0,1], and H is in the range of [0,360]. 4. the underwater target detection image enhancement method of contrast limited adaptive histogram equalization according to claim 1, is characterized in that: described step 4, The luminance vector I in the HSI space image is enhanced by contrast-limited adaptive histogram equalization algorithm, and the color information (H, S) remains unchanged. 5. the underwater target detection image enhancement method of contrast limited self-adaptive histogram equalization according to claim 1, is characterized in that: described step 5, The enhanced HSI space image is converted back to RGB space, and the converted RGB space image is represented as R'G'B', and the conversion formula is: (a) RG area (0°≤H<120°): (b) GB area (120°≤H<240°): H=H-120° (c) BR zone (240°≤H<360°): H=H-240° 6. the underwater target detection image enhancement method of contrast limited self-adaptive histogram equalization according to claim 1, is characterized in that: described step 6, Perform generalized bounded multiplication based on the adaptive gain function for the R', G', and B' components in the enhanced RGB image, and obtain the enhanced image R″G″B” based on the gradient information of the original image; generalized bounded multiplication The operation is expressed as: <mrow><mfenced open = "{" close = ""><mtable><mtr><mtd><mrow><msup><mi>R</mi><mrow><mo>&amp;prime;</mo><mo>&amp;prime;</mo></mrow></msup><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>)</mo></mrow><mo>=</mo><mi>&amp;lambda;</mi><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>)</mo></mrow><mo>&amp;CircleTimes;</mo><msup><mi>R</mi><mo>&amp;prime;</mo></msup><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>)</mo></mrow><mo>=</mo><msup><mi>&amp;phi;</mi><mrow><mo>-</mo><mn>1</mn></mrow></msup><mo>&amp;lsqb;</mo><mi>&amp;lambda;</mi><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>)</mo></mrow><msup><mi>R</mi><mo>&amp;prime;</mo></msup><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>)</mo></mrow><mo>&amp;rsqb;</mo><mo>=</mo><mfrac><mn>1</mn><mrow><msup><mrow><mo>(</mo><msup><mi>R</mi><mo>&amp;prime;</mo></msup><mo>(</mo><mrow><mi>i</mi><mo>,</mo><mi>j</mi></mrow><mo>)</mo><mo>)</mo></mrow><mrow><mi>&amp;lambda;</mi><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>)</mo></mrow></mrow></msup><mo>+</mo><mn>1</mn></mrow></mfrac></mrow></mtd></mtr><mtr><mtd><mrow><msup><mi>G</mi><mrow><mo>&amp;prime;</mo><mo>&amp;prime;</mo></mrow></msup><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>)</mo></mrow><mo>=</mo><mi>&amp;lambda;</mi><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>)</mo></mrow><mo>&amp;CircleTimes;</mo><msup><mi>G</mi><mo>&amp;prime;</mo></msup><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>)</mo></mrow><mo>=</mo><msup><mi>&amp;phi;</mi><mrow><mo>-</mo><mn>1</mn></mrow></msup><mo>&amp;lsqb;</mo><mi>&amp;lambda;</mi><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>)</mo></mrow><msup><mi>G</mi><mo>&amp;prime;</mo></msup><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>)</mo></mrow><mo>&amp;rsqb;</mo><mo>=</mo><mfrac><mn>1</mn><mrow><msup><mrow><mo>(</mo><msup><mi>G< /mi ><mo>&amp;prime;</mo></msup><mo>(</mo><mrow><mi>i</mi><mo>,</mo><mi>j</mi></mrow><mo>)</mo><mo>)</mo></mrow><mrow><mi>&amp;lambda;</mi><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>)</mo></mrow></mrow></msup><mo>+</mo><mn>1</mn></mrow></mfrac></mrow></mtd></mtr><mtr><mtd><mrow><msup><mi>B</mi><mrow><mo>&amp;prime;</mo><mo>&amp;prime;</mo></mrow></msup><mrow><mo>(</mo><mi>i</mo>mi><mo>,</mo><mi>j</mi><mo>)</mo></mrow><mo>=</mo><mi>&amp;lambda;</mi><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>)</mo></mrow><mo>&amp;CircleTimes;</mo><msup><mi>B</mi><mo>&amp;prime;</mo></msup><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>)</mo></mrow><mo>=</mo><msup><mi>&amp;phi;</mi><mrow><mo>-</mo><mn>1</mn></mrow></msup><mo>&amp;lsqb;</mo><mi>&amp;lambda;</mi><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>)</mo></mrow><msup><mi>B</mi><mo>&amp;prime;</mo></msup><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>)</mo></mrow><mo>&amp;rsqb;</mo><mo>=</mo><mfrac><mn>1</mn><mrow><msup><mrow><mo>(</mo><msup><mi>B</mi><mo>&amp;prime;</mo></msup><mo>(</mo><mrow><mi>i</mi><mo>,</mo><mi>j</mi></mrow><mo>)</mo><mo>)</mo></mrow><mrow><mi>&amp;lambda;</mi><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>)</mo></mrow></mrow></msup><mo>+</mo><mn>1</mn></mrow></mfrac></mrow></mtd></mtr></mtable></mfenced><mo>.</mo></mrow> 7. the underwater target detection image enhancement method of contrast limited self-adaptive histogram equalization according to claim 1, is characterized in that: described step 7, Transform the R, G, and B components of the enhanced image from the [0,1] range to the [0,255] range, and display the enhanced image; <mfenced open = "{" close = ""><mtable><mtr><mtd><mrow><mover><mi>R</mi><mo>~</mo></mover><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>)</mo></mrow><mo>=</mo><mn>255</mn><mo>&amp;times;</mo><msup><mi>R</mi><mrow><mo>&amp;prime;</mo><mo>&amp;prime;</mo></mrow></msup><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>)</mo></mrow></mrow></mtd></mtr><mtr><mtd><mrow><mover><mi>G</mi><mo>~</mo></mover><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>)</mo></mrow><mo>=</mo><mn>255</mn><mo>&amp;times;</mo><msup><mi>G</mi><mrow><mo>&amp;prime;</mo><mo>&amp;prime;</mo></mrow></msup><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>)</mo></mrow></mrow></mtd></mtr><mtr><mtd><mrow><mover><mi>B</mi><mo>~</mo></mover><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>)</mo></mrow><mo>=</mo><mn>255</mn><mo>&amp;times;</mo><msup><mi>B</mi><mrow><mo>&amp;prime;</mo><mo>&amp;prime;</mo></mrow></msup><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>)</mo></mrow></mrow></mtd></mtr></mtable></mfenced> The final output image can be expressed as: <mfenced open = "{" close = ""><mtable><mtr><mtd><mrow><msub><mi>R</mi><mrow><mi>o</mi><mi>u</mi><mi>t</mi></mrow></msub><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>)</mo></mrow><mo>=</mo><mfrac><mrow><mover><mi>R</mi><mo>~</mo></mover><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>)</mo></mrow><mo>-</mo><msub><mover><mi>R</mi><mo>~</mo></mover><mi>min</mi></msub></mrow><mrow><msub><mover><mi>R</mi><mo>~</mo></mover><mi>max</mi></msub><mo>-</mo><msub><mover><mi>R</mi><mo>~</mo></mover><mi>min</mi></msub></mrow></mfrac></mrow></mtd></mtr><mtr><mtd><mrow><msub><mi>G</mi><mrow><mi>o</mi><mi>u</mi><mi>t</mi></mrow></msub><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>)</mo></mrow><mo>=</mo><mfrac><mrow><mover><mi>G</mi><mo>~</mo></mover><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>)</mo></mrow><mo>-</mo><msub><mover><mi>G</mi><mo>~</mo></mover><mi>min</mi></msub></mrow><mrow><msub><mover><mi>G</mi><mo>~</mo></mover><mi>max</mi></msub><mo>-</mo><msub><mover><mi>G</mi><mo>~</mo></mover><mi>min</mi></msub></mrow></mfrac></mrow></mtd></mtr><mtr><mtd><mrow><msub><mi>B</mi><mrow><mi>o</mi><mi>u</mi><mi>t</mi></mrow></msub><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>)</mo></mrow><mo>=</mo><mfrac><mrow><mover><mi>B</mi><mo>~</mo></mover><mrow><mo>(</mo><mi>i</mi><mo>,</mo><mi>j</mi><mo>)</mo></mrow><mo>-</mo><msub><mover><mi>B</mi><mo>~</mo></mover><mi>min</mi></msub></mrow><mrow><msub><mover><mi>B</mi><mo>~</mo></mover><mi>max</mi></msub><mo>-</mo><msub><mover><mi>B</mi><mo>~</mo></mover><mi>min</mi></msub></mrow></mfrac></mrow></mtd></mtr></mtable></mfenced> in, 8. the underwater target detection image enhancement method of contrast limited self-adaptive histogram equalization according to claim 1, is characterized in that: described step eight, Quantitatively evaluate the RGB _out of the enhanced image from the aspects of mean value, contrast, information entropy and color scale, and the related quantitative evaluation index function is expressed as: mean: Wherein, μ _R , μ _G and μ _B are respectively the mean value of RGB _out three-channel color components; Contrast: In the formula, P(i,j; d,θ _k ) is the gray level co-occurrence matrix; θ _k is the angle between pixels, θ _k =(k-1)×45°, k=1,2,3,4; Information entropy: Color scale: Among them, α=RG, β=(R+G)/2-B; μ _α and μ _β are the mean values of α and β respectively, and σ _α and σ _β are the standard deviations of α and β respectively.