Imaging method and system in strong light environment based on multi-exposure fusion

Abstract

The invention provides an imaging method and system in a strong light environment based on multi-exposure fusion, and relates to the technical field of image processing, and the method comprises the steps: obtaining K RGB images of a target region, and converting the K RGB images from an RGB space to an HSV space, thereby obtaining K HSV images; extracting V channels of the K HSV images to obtain K V channel images; obtaining K weight graphs based on the K HSV images; constructing K weight Gaussian pyramids based on the K weight graphs, further constructing K V-channel Gaussian pyramids, and constructing K Laplacian pyramids based on the K V-channel Gaussian pyramids; obtaining a fused Laplacian pyramid based on the K weight Gaussian pyramids and the K Laplacian pyramids; constructing a fused image based on the fused Laplacian pyramid; and converting the fused image from the HSV space to the RGB space to obtain a final imaging result. According to the method, the calculation amount is reduced, and the fusion efficiency is greatly improved.

Description

Technical Field

[0001] The present invention relates to the field of image processing technology, and in particular to an imaging method and system in a strong light environment based on multi-exposure fusion. Background Art

[0002] Imaging technology can obtain the distribution of scene light intensity and spectrum, and contains rich scene morphology and color information. Therefore, it is widely used in key parameter detection, long-distance flight guidance, target identification and classification, equipment status monitoring and public security surveillance. Different application scenarios have different lighting conditions.

[0003] The dynamic range of a camera refers to the range between the minimum and maximum brightness values ​​of a captured image. When facing a large dynamic scene, the imaging dynamic range of an ordinary camera cannot fully cover the dynamic range of the real scene, so that it is impossible to obtain all perfect details at once. There will always be overexposure or underexposure, which can easily produce shadows and uneven lighting in the image, resulting in low image information content, loss of important information, and other problems. Therefore, in the prior art, a multi-exposure image fusion method is used to output an image with high information content.

[0004] The most widely used multi-exposure fusion method is the image fusion method based on Laplacian pyramid proposed by Mertens et al. This method decomposes the pass image into a series of sub-images with different resolutions by Gaussian convolution filtering and downsampling. Each sub-image can reflect the information of the image at some frequencies, and then performs multi-exposure fusion at different pyramid levels. Its essence is a weighted average method. The multi-exposure image fusion operation is performed on the RGB color model. Such processing will cause two major problems: first, the fusion needs to be performed on each RGB channel, which will slow down the fusion speed; second, the three color channels in the RGB model are not completely independent, which may cause color distortion and offset in the fused image. Summary of the invention

[0005] The purpose of the present invention is to provide an imaging method and system in a strong light environment based on multi-exposure fusion, which reduces the amount of calculation and greatly improves the fusion efficiency.

[0006] An imaging method in a strong light environment based on multi-exposure fusion, comprising:

[0007] Acquire K RGB images of the target area; the K RGB images have different exposure times, and K is a positive integer greater than 1;

[0008] Convert the K RGB images from the RGB space to the HSV space to obtain K HSV images;

[0009] Extracting the V channels of the K HSV images to obtain K V channel images;

[0010] Based on the K HSV images, we get K weight maps:

[0011] Based on the K HSV images, K contrast images are obtained; the expression is:

[0012] C k (i,j)=|4Y k (i,j)-Y k (i+1,j)-Y k (i-1,j)-Y k (i,j+1)-Y k (i,j-1)|;

[0013] Where: Y k (i, j) represents the pixel value of the kth HSV image at the pixel point (i, j), k = [1, 2, ..., K], C k (i, j) represents the pixel value of the k-th contrast image at the pixel point (i, j);

[0014] Based on the K HSV images, K saturation images are obtained; the expression is:

[0015]

[0016] In the formula, represents the pixel value of the kth saturation image at pixel (i, j), S k (i, j) represents the pixel value of the S channel of the kth HSV image at the pixel point (i, j), u represents the parameter value, 0<u<1;

[0017] Multiply the K contrast images and the K saturation images to obtain the K weight maps; the expression is:

[0018]

[0019] Where: W k (i, j) represents the pixel value of the kth weighted image at the pixel point (i, j), W c represents the contrast index, W s represents the saturation index;

[0020] K weight Gaussian pyramids are constructed based on the K weight maps, K V-channel Gaussian pyramids are constructed based on the K V-channel images, and K Laplacian pyramids are constructed based on the K V-channel Gaussian pyramids; the weight Gaussian pyramid has M-1 layers, the V-channel Gaussian pyramid has M layers, and the Laplacian pyramid has M-1 layers;

[0021] A fused Laplacian pyramid is obtained based on the K weighted Gaussian pyramids and the K Laplacian pyramids:

[0022] Multiplying the mth layer of the kth weighted Gaussian pyramid by the mth layer of K Laplacian pyramids to obtain a kth mth layer initial fused image, superimposing the K mth layer initial fused images to obtain an mth layer fused image, performing this process on each layer to obtain the fused Laplacian pyramid;

[0023] Constructing a fused image based on the fused Laplacian pyramid;

[0024] The fused image is converted from the HSV space to the RGB space to obtain the final imaging result.

[0025] Optionally, the K RGB images are converted from the RGB space to the HSV space to obtain K HSV images, and the expression is:

[0026]

[0027] V=C max ;

[0028] Where: H represents the pixel value of the HSV space H channel, S represents the pixel value of the HSV space S channel, V represents the pixel value of the HSV space channel, R'=R / 255, G'=G / 255, B'=B / 255,

[0029] C max =max(R',G',B'),C min =min(R',G',B'),Δ=C max -C min , R represents the pixel value of the R channel in the RGB space, G represents the pixel value of the G channel in the RGB space, and B represents the pixel value of the B channel in the RGB space.

[0030] Optionally, the fused image is converted from the HSV space to the RGB space to obtain a final imaging result, which is expressed as:

[0031]

[0032] (R, G, B) = ((R'+m)×255, (G'+m)×255, (B'+m)×255);

[0033] Among them, when 0≤H<360, 0≤S<1, 0≤V<1, C=V×S, X=C×(1-|(H / 60°)mod2-1), m=VC.

[0034] Optionally, constructing a fused image based on the fused Laplacian pyramid is specifically:

[0035] The M-1th layer of the fused Laplacian pyramid is upsampled and superimposed with the M-2th layer of the fused Laplacian pyramid to obtain an M-2th layer superimposed image, the M-2th layer superimposed image is upsampled and superimposed with the M-3th layer of the fused Laplacian pyramid to obtain an M-3th layer superimposed image, and this process is repeated until a second layer superimposed image is obtained, and the second layer superimposed image is superimposed with the first layer of the fused Laplacian pyramid to obtain the fused image.

[0036] The present invention also provides an imaging system in a strong light environment based on multi-exposure fusion, which comprises:

[0037] A data acquisition module, used for acquiring K RGB images of a target area; the K RGB images have different exposure times, and K is a positive integer greater than 1;

[0038] A first conversion module, used for converting the K RGB images from the RGB space to the HSV space to obtain K HSV images;

[0039] An image extraction module, used for extracting the V channels of the K HSV images to obtain K V channel images;

[0040] The weighted image module obtains K weighted images based on the K HSV images:

[0041] A contrast unit is used to obtain K contrast images based on the K HSV images; the expression is:

[0042] C k (i,j)=|4Y k (i,j)-Y k (i+1,j)-Y k (i-1,j)-Y k (i,j+1)-Y k (i,j-1)|;

[0043] Where: Y k (i, j) represents the pixel value of the kth HSV image at the pixel point (i, j), k = [1, 2, ..., K], C k (i, j) represents the pixel value of the k-th contrast image at the pixel point (i, j);

[0044] The saturation unit is used to obtain K saturation images based on the K HSV images; the expression is:

[0045]

[0046] In the formula, represents the pixel value of the kth saturation image at pixel (i, j), S k (i, j) represents the pixel value of the S channel of the kth HSV image at the pixel point (i, j), u represents the parameter value, 0<u<1;

[0047] The weight image unit is used to multiply the K contrast images with the K saturation images to obtain the K weight images; the expression is:

[0048]

[0049] Where: W k (i, j) represents the pixel value of the kth weighted image at the pixel point (i, j), W c represents the contrast index, W s represents the saturation index;

[0050] A pyramid module is used to construct K weight Gaussian pyramids based on the K weight maps, construct K V-channel Gaussian pyramids based on the K V-channel images, and construct K Laplacian pyramids based on the K V-channel Gaussian pyramids; the weight Gaussian pyramid has M-1 layers, the V-channel Gaussian pyramid has M layers, and the Laplacian pyramid has M-1 layers;

[0051] A fusion module is used to obtain a fused Laplacian pyramid based on the K weighted Gaussian pyramids and the K Laplacian pyramids:

[0052] Multiplying the mth layer of the kth weighted Gaussian pyramid by the mth layer of K Laplacian pyramids to obtain a kth mth layer initial fused image, superimposing the K mth layer initial fused images to obtain an mth layer fused image, performing this process on each layer to obtain the fused Laplacian pyramid;

[0053] An image reconstruction module, used for constructing a fused image based on the fused Laplacian pyramid;

[0054] The second conversion module is used to convert the fused image from the HSV space to the RGB space to obtain a final imaging result.

[0055] Optionally, the first conversion module expression is:

[0056]

[0057] V=C max ;

[0058] Where: H represents the pixel value of the HSV space H channel, S represents the pixel value of the S HSV space S channel, V represents the pixel value of the HSV space channel, R'=R / 255, G'=G / 255, B'=B / 255, C max =max(R',G',B'),C min =min(R',G',B'),Δ=C max -C min , R represents the pixel value of the R channel in the RGB space, G represents the pixel value of the G channel in the RGB space, and B represents the pixel value of the B channel in the RGB space.

[0059] Optionally, the second conversion module expression is:

[0060]

[0061] (R, G, B) = ((R'+m)×255, (G'+m)×255, (B'+m)×255);

[0062] Among them, when 0≤H<360, 0≤S<1, 0≤V<1, C=V×S, X=C×(1-|(H / 60°)mod2-1), m=VC.

[0063] Optionally, the image reconstruction module is specifically:

[0064] The M-1th layer of the fused Laplacian pyramid is upsampled and superimposed with the M-2th layer of the fused Laplacian pyramid to obtain an M-2th layer superimposed image, the M-2th layer superimposed image is upsampled and superimposed with the M-3th layer of the fused Laplacian pyramid to obtain an M-3th layer superimposed image, and this process is repeated until a second layer superimposed image is obtained, and the second layer superimposed image is superimposed with the first layer of the fused Laplacian pyramid to obtain the fused image.

[0065] The effects of the present invention are as follows:

[0066] The imaging method in a strong light environment based on multi-exposure fusion in the present invention first converts the image in the RGB color space into the HSV color space, thereby solving the problems of color shift and large amount of calculation in the RGB color space. Then, only the V component in the HSV image is weightedly fused. Finally, the fused image is converted back to the RGB color space to obtain the final fused image. Under the premise of ensuring the fusion effect, the amount of calculation is reduced and the fusion efficiency is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] Figure 1 It is a flow chart of the imaging method in a strong light environment based on multi-exposure fusion of the present invention. DETAILED DESCRIPTION

[0068] Hereinafter, embodiments of the present invention will be described with reference to the drawings.

[0069] Figure 1 This is a flow chart of the imaging method in a strong light environment based on multi-exposure fusion of the present invention. Figure 1 As shown, the present invention provides an imaging method in a strong light environment based on multi-exposure fusion, which includes:

[0070] S1, obtaining K RGB images of the target area. The exposure times of the K RGB images are different, and K is a positive integer greater than 1.

[0071] S2, convert K RGB images from RGB space to HSV space to obtain K HSV images.

[0072] Preferably, S2 is expressed as:

[0073]

[0074] V=C max ;

[0075] Where: H represents the pixel value of the HSV space H channel, S represents the pixel value of the HSV space S channel, V represents the pixel value of the HSV space channel, R'=R / 255, G'=G / 255, B'=B / 255,

[0076] C max =max(R',G',B'),C min =min(R',G',B'),Δ=C max -C min , R represents the pixel value of the R channel in the RGB space, G represents the pixel value of the G channel in the RGB space, and B represents the pixel value of the B channel in the RGB space.

[0077] S3, extracting the V channels of the K HSV images to obtain K V channel images.

[0078] According to the definition of image dynamic range, the brightness of the image determines its dynamic range. HSV space is a widely used color space, in which H represents hue, S represents saturation, and V represents brightness value. Therefore, the present invention mainly operates on the V component in the multi-exposure fusion process. Since the fusion process is only performed on the V component, the operating efficiency is greatly improved compared to the traditional fusion method in the RGB space, and only processing the V component will not affect the color of the image, which can solve the problem of color shift to a certain extent.

[0079] S4, obtain K weight maps based on the K HSV images.

[0080] Specifically, S4 includes:

[0081] S41, obtain K contrast images based on K HSV images. The expression is:

[0082] C k (i,j)=|4Y k (i,j)-Y k (i+1,j)-Y k (i-1,j)-Y k (i,j+1)-Y k (i,j-1)|;

[0083] Where: Y k (i, j) represents the pixel value of the kth HSV image at the pixel point (i, j), k = [1, 2, ..., K], C k (i, j) represents the pixel value of the k-th contrast image at the pixel point (i, j).

[0084] When processing multi-exposure images, a Laplacian convolution filter is used to process the image, and the absolute value of the Laplacian filter response value is used as the contrast weight of each area of ​​the image.

[0085] S42, obtain K saturation images based on K HSV images. The expression is:

[0086]

[0087] In the formula, represents the pixel value of the kth saturation image at pixel (i, j), S k (i, j) represents the pixel value of the S channel of the kth HSV image at the pixel point (i, j), u represents the parameter value, 0<u<1.

[0088] In the HSV color space, saturation (hue) determines the degree of proximity between a color and the purest spectral color. Hue is usually considered to be a mixture of a specific spectral color and white. The higher the proportion in the spectral color, the greater the saturation, the closer the color appears to the pure spectral color, and the deeper and brighter the color appears. The standard value of saturation is generally between 0 and 1. An increase in this value means that the color is more saturated. In the process of saturation factor fusion, in order to avoid the obtained weight being 0, an extremely small value u is introduced.

[0089] S43, multiplying the K contrast images with the K saturation images to obtain K weight maps. The expression is:

[0090]

[0091] Where: Wk (i, j) represents the pixel value of the kth weighted image at the pixel point (i, j), W c Represents the contrast index, W s Represents the saturation index.

[0092] S5: construct K weight Gaussian pyramids based on K weight images, construct K V-channel Gaussian pyramids based on K V-channel images, and construct K Laplacian pyramids based on K V-channel Gaussian pyramids. The weight Gaussian pyramid has M-1 layers, the V-channel Gaussian pyramid has M layers, and the Laplacian pyramid has M-1 layers.

[0093] Specifically, a V-channel Gaussian pyramid is constructed based on the V-channel image, specifically:

[0094] The V channel image is used as the first layer of the V channel Gaussian pyramid, and the V channel image is convoluted and filtered based on the Gaussian kernel, and then downsampled to obtain the second layer image, and the second layer image is used as the second layer of the V channel Gaussian pyramid, and the second layer image is convoluted and filtered based on the Gaussian kernel, and then downsampled to obtain the third layer image, and the third layer image is used as the third layer of the V channel Gaussian pyramid, and this process is repeated until the M layer image is obtained, and the M layer image is used as the M layer of the V channel Gaussian pyramid. Downsampling is to remove even rows and columns.

[0095] Construct a Laplacian pyramid based on the V-channel Gaussian pyramid, specifically:

[0096] The M-th layer of the V-channel Gaussian pyramid is upsampled and then subtracted from the M-1th layer of the V-channel Gaussian pyramid to obtain the M-1th layer of the Laplacian pyramid. The M-1th layer of the V-channel Gaussian pyramid is upsampled and then subtracted from the M-2th layer of the V-channel Gaussian pyramid to obtain the M-2th layer of the Laplacian pyramid. This process is repeated until the first layer of the Laplacian pyramid is obtained.

[0097] S6, obtaining a fused Laplacian pyramid based on the K weighted Gaussian pyramids and the K Laplacian pyramids.

[0098] S6 is specifically:

[0099] The mth layer of the kth weighted Gaussian pyramid is multiplied by the mth layer of the K Laplacian pyramids to obtain the kth mth layer initial fused image, and the K mth layer initial fused images are superimposed to obtain the mth layer fused image. This process is performed on each layer to obtain a fused Laplacian pyramid.

[0100] S7, constructs a fused image based on the fused Laplacian pyramid.

[0101] Furthermore, S7 is specifically:

[0102] The M-1th layer of the fused Laplacian pyramid is upsampled and superimposed with the M-2th layer of the fused Laplacian pyramid to obtain the M-2th layer superimposed image, the M-2th layer superimposed image is upsampled and superimposed with the M-3th layer of the fused Laplacian pyramid to obtain the M-3th layer superimposed image, and this process is repeated until the second layer superimposed image is obtained, and the second layer superimposed image is superimposed with the first layer of the fused Laplacian pyramid to obtain a fused image.

[0103] S8, converting the fused image from the HSV space to the RGB space to obtain the final imaging result.

[0104] Preferably, the S8 expression is:

[0105]

[0106] (R, G, B) = ((R'+m)×255, (G'+m)×255, (B'+m)×255);

[0107] Among them, when 0≤H<360, 0≤S<1, 0≤V<1, C=V×S, X=C×(1-|(H / 60°)mod2-1), m=VC, mod represents the remainder after dividing two values, that is, the remainder after (H / 60) is divided by 2.

[0108] The present invention also provides an imaging system in a strong light environment based on multi-exposure fusion, which comprises:

[0109] The data acquisition module is used to acquire K RGB images of the target area. The exposure times of the K RGB images are different, and K is a positive integer greater than 1.

[0110] The first conversion module is used to convert K RGB images from RGB space to HSV space to obtain K HSV images.

[0111] The image extraction module is used to extract the V channels of K HSV images to obtain K V channel images.

[0112] The weighted image module obtains K weighted images based on K HSV images:

[0113] The contrast unit is used to obtain K contrast images based on K HSV images. The expression is:

[0114] C k (i,j)=|4Y k (i,j)-Y k (i+1,j)-Y k (i-1,j)-Y k (i,j+1)-Y k(i,j-1)|;

[0115] Where: Y k (i, j) represents the pixel value of the kth HSV image at the pixel point (i, j), k = [1, 2, ..., K], C k (i, j) represents the pixel value of the k-th contrast image at the pixel point (i, j).

[0116] The saturation unit is used to obtain K saturation images based on K HSV images. The expression is:

[0117]

[0118] In the formula, represents the pixel value of the kth saturation image at pixel (i, j), S k (i, j) represents the pixel value of the S channel of the kth HSV image at the pixel point (i, j), u represents the parameter value, 0<u<1.

[0119] The weight image unit is used to multiply K contrast images with K saturation images to obtain K weight images. The expression is:

[0120]

[0121] Where: W k (i, j) represents the pixel value of the kth weighted image at the pixel point (i, j), W c Represents the contrast index, W s Represents the saturation index.

[0122] The pyramid module is used to construct K weight Gaussian pyramids based on K weight images, K V-channel Gaussian pyramids based on K V-channel images, and K Laplacian pyramids based on K V-channel Gaussian pyramids. The weight Gaussian pyramid has M-1 layers, the V-channel Gaussian pyramid has M layers, and the Laplacian pyramid has M-1 layers.

[0123] The fusion module is used to obtain a fused Laplacian pyramid based on K weighted Gaussian pyramids and K Laplacian pyramids:

[0124] The mth layer of the kth weighted Gaussian pyramid is multiplied by the mth layer of the K Laplacian pyramids to obtain the kth mth layer initial fused image, and the K mth layer initial fused images are superimposed to obtain the mth layer fused image. This process is performed on each layer to obtain a fused Laplacian pyramid.

[0125] Image reconstruction module, used to construct a fused image based on the fused Laplacian pyramid.

[0126] The second conversion module is used to convert the fused image from the HSV space to the RGB space to obtain the final imaging result.

[0127] Preferably, the first conversion module expression is:

[0128]

[0129] V=C max ;

[0130] Where: H represents the pixel value of the HSV space H channel, S represents the pixel value of the S HSV space S channel, V represents the pixel value of the HSV space channel, R'=R / 255, G'=G / 255, B'=B / 255, C max =max(R',G',B'),C min =min(R',G',B'),Δ=C max -C min , R represents the pixel value of the R channel in the RGB space, G represents the pixel value of the G channel in the RGB space, and B represents the pixel value of the B channel in the RGB space.

[0131] Preferably, the second conversion module expression is:

[0132]

[0133] (R, G, B) = ((R'+m)×255, (G'+m)×255, (B'+m)×255);

[0134] Among them, when 0≤H<360, 0≤S<1, 0≤V<1, C=V×S, X=C×(1-|(H / 60°)mod2-1), m=VC.

[0135] Preferably, the image reconstruction module is specifically:

[0136] The M-1th layer of the fused Laplacian pyramid is upsampled and superimposed with the M-2th layer of the fused Laplacian pyramid to obtain the M-2th layer superimposed image, the M-2th layer superimposed image is upsampled and superimposed with the M-3th layer of the fused Laplacian pyramid to obtain the M-3th layer superimposed image, and this process is repeated until the second layer superimposed image is obtained, and the second layer superimposed image is superimposed with the first layer of the fused Laplacian pyramid to obtain a fused image.

[0137] The embodiments described above are only descriptions of the preferred implementation modes of the present invention, and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should all fall within the protection scope determined by the claims of the present invention.

Claims

1. An imaging method in a strong light environment based on multi-exposure fusion, characterized in that: It includes: Get K RGB images of the target area; The exposure times of the K RGB images are different, and K is a positive integer greater than 1; Convert the K RGB images from the RGB space to the HSV space to obtain K HSV images; Extracting the V channels of the K HSV images to obtain K V channel images; Based on the K HSV images, we get K weight maps: Based on the K HSV images, K contrast images are obtained; the expression is: C k (i,j)=|4Y k (i,j)-Y k (i+1,j)-Y k (i-1,j)-Y k (i,j+1)-Y k (i,j-1)|; Where: Y k (i, j) represents the pixel value of the kth HSV image at the pixel point (i, j), k = [1, 2, ..., K], C k (i, j) represents the pixel value of the k-th contrast image at the pixel point (i, j); Based on the K HSV images, K saturation images are obtained; the expression is: In the formula, represents the pixel value of the kth saturation image at pixel (i, j), S k (i, j) represents the pixel value of the S channel of the kth HSV image at the pixel point (i, j), u represents the parameter value, 0<u<1; Multiply the K contrast images and the K saturation images to obtain the K weight maps; the expression is: Where: W k (i, j) represents the pixel value of the kth weighted image at the pixel point (i, j), W c Represents the contrast index, W s represents the saturation index; K weight Gaussian pyramids are constructed based on the K weight maps, K V-channel Gaussian pyramids are constructed based on the K V-channel images, and K Laplacian pyramids are constructed based on the K V-channel Gaussian pyramids; the weight Gaussian pyramid has M-1 layers, the V-channel Gaussian pyramid has M layers, and the Laplacian pyramid has M-1 layers; A fused Laplacian pyramid is obtained based on the K weighted Gaussian pyramids and the K Laplacian pyramids: Multiplying the mth layer of the kth weighted Gaussian pyramid by the mth layer of K Laplacian pyramids to obtain a kth mth layer initial fused image, superimposing the K mth layer initial fused images to obtain an mth layer fused image, performing this process on each layer to obtain the fused Laplacian pyramid; Constructing a fused image based on the fused Laplacian pyramid; The fused image is converted from the HSV space to the RGB space to obtain the final imaging result.

2. The imaging method in a strong light environment based on multi-exposure fusion according to claim 1, characterized in that: The K RGB images are converted from the RGB space to the HSV space to obtain K HSV images, and the expression is: V=C max ; Where: H represents the pixel value of the HSV space H channel, S represents the pixel value of the S HSV space S channel, V represents the pixel value of the HSV space channel, R'=R / 255, G'=G / 255, B'=B / 255, C max =max(R',G',B'),C min =min(R',G',B'),Δ=C max -C min , R represents the pixel value of the R channel in the RGB space, G represents the pixel value of the G channel in the RGB space, and B represents the pixel value of the B channel in the RGB space.

3. The imaging method in a strong light environment based on multi-exposure fusion according to claim 1, characterized in that: The fused image is converted from the HSV space to the RGB space to obtain the final imaging result, which is expressed as: (R, G, B) = ((R'+m)×255, (G'+m)×255, (B'+m)×255); Among them, when 0≤H<360, 0≤S<1, 0≤V<1, C=V×S, X=C×(1-|(H / 60°)mod2-1), m=VC.

4. The imaging method in a strong light environment based on multi-exposure fusion according to claim 1, characterized in that: The constructing of the fused image based on the fused Laplacian pyramid is specifically as follows: The M-1th layer of the fused Laplacian pyramid is upsampled and superimposed with the M-2th layer of the fused Laplacian pyramid to obtain an M-2th layer superimposed image, the M-2th layer superimposed image is upsampled and superimposed with the M-3th layer of the fused Laplacian pyramid to obtain an M-3th layer superimposed image, and this process is repeated until a second layer superimposed image is obtained, and the second layer superimposed image is superimposed with the first layer of the fused Laplacian pyramid to obtain the fused image.

5. An imaging system in a strong light environment based on multi-exposure fusion, characterized in that: It includes: A data acquisition module is used to acquire K RGB images of the target area; The exposure times of the K RGB images are different, and K is a positive integer greater than 1; A first conversion module, used for converting the K RGB images from the RGB space to the HSV space to obtain K HSV images; An image extraction module, used for extracting the V channels of the K HSV images to obtain K V channel images; The weighted image module obtains K weighted images based on the K HSV images: A contrast unit is used to obtain K contrast images based on the K HSV images; the expression is: C k (i,j)=|4Y k (i,j)-Y k (i+1,j)-Y k (i-1,j)-Y k (i,j+1)-Y k (i,j-1)|; Where: Y k (i, j) represents the pixel value of the kth HSV image at the pixel point (i, j), k = [1, 2, ..., K], C k (i, j) represents the pixel value of the k-th contrast image at the pixel point (i, j); The saturation unit is used to obtain K saturation images based on the K HSV images; the expression is: In the formula, represents the pixel value of the kth saturation image at pixel (i, j), S k (i, j) represents the pixel value of the S channel of the kth HSV image at the pixel point (i, j), u represents the parameter value, 0<u<1; The weight image unit is used to multiply the K contrast images with the K saturation images to obtain the K weight images; the expression is: Where: W k (i, j) represents the pixel value of the kth weighted image at the pixel point (i, j), W c Represents the contrast index, W s represents the saturation index; A pyramid module is used to construct K weight Gaussian pyramids based on the K weight maps, construct K V-channel Gaussian pyramids based on the K V-channel images, and construct K Laplacian pyramids based on the K V-channel Gaussian pyramids; the weight Gaussian pyramid has M-1 layers, the V-channel Gaussian pyramid has M layers, and the Laplacian pyramid has M-1 layers; A fusion module is used to obtain a fused Laplacian pyramid based on the K weighted Gaussian pyramids and the K Laplacian pyramids: Multiplying the mth layer of the kth weighted Gaussian pyramid by the mth layer of K Laplacian pyramids to obtain a kth mth layer initial fused image, superimposing the K mth layer initial fused images to obtain an mth layer fused image, performing this process on each layer to obtain the fused Laplacian pyramid; An image reconstruction module, used for constructing a fused image based on the fused Laplacian pyramid; The second conversion module is used to convert the fused image from the HSV space to the RGB space to obtain a final imaging result.

6. The imaging system based on multi-exposure fusion in a strong light environment according to claim 5, characterized in that: The first conversion module expression is: V=C max ; Where: H represents the pixel value of the HSV space H channel, S represents the pixel value of the S HSV space S channel, V represents the pixel value of the HSV space channel, R'=R / 255, G'=G / 255, B'=B / 255, C max =max(R',G',B'),C min =min(R',G',B'),Δ=C max -C min , R represents the pixel value of the R channel in the RGB space, G represents the pixel value of the G channel in the RGB space, and B represents the pixel value of the B channel in the RGB space.

7. The imaging system based on multi-exposure fusion in a strong light environment according to claim 5, characterized in that: The second conversion module expression is: (R, G, B) = ((R'+m)×255, (G'+m)×255, (B'+m)×255); Among them, when 0≤H<360, 0≤S<1, 0≤V<1, C=V×S, X=C×(1-|(H / 60°)mod2-1), m=VC.

8. The imaging system based on multi-exposure fusion in a strong light environment according to claim 5, characterized in that: The image reconstruction module is specifically: The M-1th layer of the fused Laplacian pyramid is upsampled and superimposed with the M-2th layer of the fused Laplacian pyramid to obtain an M-2th layer superimposed image, the M-2th layer superimposed image is upsampled and superimposed with the M-3th layer of the fused Laplacian pyramid to obtain an M-3th layer superimposed image, and this process is repeated until a second layer superimposed image is obtained, and the second layer superimposed image is superimposed with the first layer of the fused Laplacian pyramid to obtain the fused image.

Images