Remote sensing image defogging method and device based on YUV spatial feature guidance

Abstract

The invention discloses a remote sensing image defogging method and device based on YUV spatial feature guidance, and the method comprises the following steps: disassembling a haze image, and obtaining a brightness component and a chrominance component corresponding to the haze image; carrying out defogging processing on the brightness component, and carrying out enhancement processing on the chrominance component; and splicing and restoring the defogged brightness component and the enhanced chrominance component to obtain a defogged image. The defogging effect of the remote sensing image can be improved.

Description

Technical Field

[0001] The present invention relates to the field of image processing, and in particular to a remote sensing image defogging method and device based on YUV spatial feature guidance. Background Art

[0002] With the rapid development of remote sensing technology, the demand for image processing in earth observation is increasing, and remote sensing image processing in haze environment has gradually become an important research direction in this field. Remote sensing images are affected by the complex imaging environment and atmospheric conditions. Especially under weather conditions such as haze and dust, images often show reduced contrast, blurred details, and distorted colors. These problems not only significantly reduce the availability of remote sensing data, but also challenge its accuracy and efficiency in applications such as environmental monitoring, disaster assessment, and agricultural resource management.

[0003] The quality degradation of remote sensing images is closely related to atmospheric scattering and absorption. First, the scattering of light by suspended particles in a haze environment causes loss of image details and foggy blur. Second, the multipath effect of scattering introduces an increase in global brightness, which significantly reduces image contrast. At the same time, the absorption and scattering characteristics of light vary with wavelength, resulting in significant deviations between the color performance of the image and the actual scene, such as gray or white. In addition, the complex distribution of media will also disturb the propagation path of light, making the degradation characteristics of remote sensing images show significant non-uniformity. These problems not only pose great challenges to traditional image processing algorithms, but also put forward higher requirements for the cross-scene stability and generalization performance of deep learning algorithms. In order to solve the quality degradation problem of haze remote sensing images, researchers have proposed a variety of image enhancement methods, which are mainly divided into algorithms based on physical models and deep learning methods based on data-driven. The physical model method is based on the atmospheric scattering model and restores image contrast and details by estimating transmittance and atmospheric light parameters. This type of method has a solid theoretical foundation, but is sensitive to parameter settings and has limited performance in complex scenes. Deep learning methods use a data-driven approach to learn the mapping relationship between image degradation and clear images through a large number of training samples. They perform well in specific scenarios, but have high requirements for computing resources and are prone to instability in unprecedented scenarios. Summary of the invention

[0004] In order to overcome the above-mentioned defects of the prior art, the present invention provides a remote sensing image defogging method and device based on YUV spatial feature guidance, which can improve the defogging effect of remote sensing images.

[0005] An embodiment of the present invention provides a remote sensing image defogging method based on YUV spatial feature guidance, comprising the following steps: Decompose the haze image to obtain the brightness component and chromaticity component corresponding to the haze image; Performing a defogging process on the brightness component and an enhancement process on the chrominance component; The defogging brightness component and the enhanced chrominance component are stitched and restored to obtain a defogging image.

[0006] Furthermore, the decomposing of the haze image to obtain the brightness component and the chrominance component corresponding to the haze image specifically includes: The haze image is mapped from the RGB space to the YUV space to obtain a YUV haze image; wherein the mapping relationship from the RGB space to the YUV space is specifically: ; The YUV haze image is dimensionally decomposed to obtain the brightness component and the chrominance component; wherein the brightness component is the Y channel component corresponding to the YUV haze image, and the chrominance component is the U channel component and the V channel component corresponding to the YUV haze image.

[0007] Furthermore, before performing defogging processing on the brightness component, the method further includes: calculating the ambient light value and scene transmittance corresponding to the haze image, and the specific calculation process includes: Mapping the haze image to a grayscale space to obtain a grayscale image; From the grayscale image dark channel brightness b ‰ pixels, determine the pixel value of the pixel with the largest corresponding brightness value as the ambient light value; wherein, b is the preset ratio value; The scene transmittance is calculated according to the ambient light value and a preset atmospheric scattering model; wherein the calculation formula of the scene transmittance is specifically as follows:

[0008] in, is the scene transmittance, is the ambient light value, is the preset correction factor, is the preset atmospheric scattering model, To preset a clear image, The pixel point in the grayscale image A preset square area centered on the

[0009] Furthermore, the defogging of the brightness component specifically includes: According to the ambient light value and the scene transmittance, the brightness component is defogged to obtain a clear brightness component; wherein the specific calculation formula of the defogging process is:

[0010] in, is the clear brightness component, is the brightness component, is the preset stability constant.

[0011] Furthermore, the enhancing process of the chrominance component specifically includes: According to the ambient light value and the scene transmittance, the chromaticity component is enhanced to obtain a clear U channel component and a clear V channel component; wherein the specific calculation formula of the enhancement processing is:

[0012]

[0013] in, , are the clear U channel component and the clear V channel component respectively, , are respectively the U channel component and the V channel component in the chrominance component, is the preset stability constant.

[0014] Furthermore, the restoration of the defogging brightness component and the enhanced chrominance component to obtain a defogging image specifically includes: According to the defogging brightness component and the enhanced chrominance component, a defogging YUV image is obtained by dimension stitching; The defogging YUV image is mapped from the YUV space to the RGB space to obtain the defogging image; wherein the mapping relationship from the YUV space to the RGB space is specifically: .

[0015] Another embodiment of the present invention provides a remote sensing image defogging device based on YUV spatial feature guidance, including: a disassembly module, a processing module and a restoration module; The disassembly module is used to disassemble the haze image to obtain the brightness component and chromaticity component corresponding to the haze image; The processing module is used to perform a defogging process on the brightness component and an enhancement process on the chrominance component; The restoration module is used to restore the defogging brightness component and the enhanced chrominance component to obtain a defogging image.

[0016] Furthermore, the decomposition module is used to decompose the haze image to obtain the brightness component and chrominance component corresponding to the haze image, specifically including: The haze image is mapped from the RGB space to the YUV space to obtain a YUV haze image; wherein the mapping relationship from the RGB space to the YUV space is specifically: ; The YUV haze image is dimensionally decomposed to obtain the brightness component and the chrominance component; wherein the brightness component is the Y channel component corresponding to the YUV haze image, and the chrominance component is the U channel component and the V channel component corresponding to the YUV haze image.

[0017] Furthermore, before the processing module performs defogging on the brightness component, it also includes: calculating the ambient light value and scene transmittance corresponding to the haze image, and the specific calculation process includes: Mapping the haze image to a grayscale space to obtain a grayscale image; From the grayscale image dark channel brightness b ‰ pixels, determine the pixel value of the pixel with the largest corresponding brightness value as the ambient light value; wherein, b is the preset ratio value; The scene transmittance is calculated according to the ambient light value and a preset atmospheric scattering model; wherein the calculation formula of the scene transmittance is specifically as follows:

[0018] in, is the scene transmittance, is the ambient light value, is the preset correction factor, is the preset atmospheric scattering model, The pixel point in the grayscale image A preset square area centered on the

[0019] Furthermore, the restoration module is used to stitch and restore the defogged brightness component and the enhanced chrominance component to obtain a defogged image, which specifically includes: According to the defogging brightness component and the enhanced chrominance component, a defogging YUV image is obtained by dimension stitching; The defogging YUV image is mapped from the YUV space to the RGB space to obtain the defogging image; wherein the mapping relationship from the YUV space to the RGB space is specifically: .

[0020] Compared with the prior art, the beneficial effects of the present invention are: By designing restoration strategies for the brightness and chromaticity components of haze images respectively, it is possible to effectively restore the overall clarity and detail level of the image while regulating and enhancing the chromaticity information, repairing the color shift and saturation reduction problems caused by haze or light scattering, and better solving the problems of visual blur and color degradation in remote sensing haze images, thereby improving the dehazing effect of haze images. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 A schematic flow chart of a remote sensing image defogging method based on YUV spatial feature guidance provided by an embodiment of the present invention.

[0022] Figure 2 A qualitative and quantitative comparison diagram of a clear remote sensing image and a haze remote sensing image in RGB and YUV space provided by an embodiment of the present invention.

[0023] Figure 3 A quantitative difference comparison diagram of the histogram distribution of a clear remote sensing image and a haze remote sensing image in YUV space is provided in one embodiment of the present invention.

[0024] Figure 4 A schematic diagram of visual representation of each channel in a YUV color space corresponding to a processed remote sensing image provided by an embodiment of the present invention.

[0025] Figure 5 A schematic diagram of defogging results of a remote sensing image defogging method based on YUV spatial feature guidance provided by an embodiment of the present invention.

[0026] Figure 6 A schematic structural diagram of a remote sensing image defogging device based on YUV spatial feature guidance is provided in another embodiment of the present invention. DETAILED DESCRIPTION

[0027] The drawings are for illustrative purposes only and should not be construed as limiting the present patent; It is understandable to those skilled in the art that some well-known structures and their descriptions may be omitted in the drawings.

[0028] The technical solution of the present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0029] Reference Figure 1 , is a flow chart of a remote sensing image defogging method based on YUV spatial feature guidance provided by an embodiment of the present invention, comprising the following steps: S1: Decompose the haze image to obtain the brightness component and chromaticity component corresponding to the haze image; S2: performing a defogging process on the brightness component and an enhancement process on the chrominance component; S3: Restoring the defogging brightness component and the enhanced chrominance component by splicing to obtain a defogging image.

[0030] For step S1, specifically, the decomposing of the haze image to obtain the brightness component and chrominance component corresponding to the haze image specifically includes: The haze image is mapped from the RGB space to the YUV space to obtain a YUV haze image; wherein the mapping relationship from the RGB space to the YUV space is specifically: ; The YUV haze image is dimensionally decomposed to obtain the brightness component and the chrominance component; wherein the brightness component is the Y channel component corresponding to the YUV haze image, and the chrominance component is the U channel component and the V channel component corresponding to the YUV haze image.

[0031] In a preferred embodiment, Peak Signal-to-Noise Ratio (PSNR) and Structural SIMilarity (SSIM) are two important indicators for evaluating image quality, which are widely used in the field of image processing to evaluate the effect of image reconstruction or enhancement. PSNR is an indicator that measures the difference between two images (usually the original image and the reconstructed image), reflecting the quality of image reconstruction. PSNR is expressed in decibels (dB). The larger the value, the better the image quality. The corresponding calculation formula is as follows:

[0032] in, MAX Indicates the maximum possible value of the image pixel value. For example, for an 8-bit image, MAX = 255; MSE is the mean square error, which represents the square mean of the corresponding pixel differences between the two images.

[0033] SSIM is an image quality evaluation index based on human perception, focusing on the structural information and visual quality of the image. It simulates the human eye's perception mechanism of image brightness, contrast and structure, and its value range is [-1, 1], where 1 means that the two images are exactly the same, which can be expressed as:

[0034] in, x and y Represents two contrasting images; μ x and μ y Respectively represent x and yThe average value of represents brightness; σ x and σ y Respectively represent x and y The variance of , which indicates the contrast; σ xy for x and y The covariance of , indicating structural similarity; C 1 and C 2 represents a constant to maintain computational stability.

[0035] The channel characteristics of clear remote sensing images and haze remote sensing images in YUV space are disassembled, and the PSNR and SSIM indicators are used to evaluate the pixel and structural interference caused by haze on the three channel components of Y, U, and V. The smaller the PSNR and SSIM values, the more serious the image degradation. Figure 2 , which is a qualitative and quantitative comparison diagram of a clear remote sensing image and a haze remote sensing image in RGB and YUV space provided by an embodiment of the present invention, wherein the first row is the clear remote sensing image feature, and the last row is the corresponding spatial feature of the remote sensing haze image. Figure 2 It can be seen that, different from the RGB spatial features, the remote sensing haze image has different degrees of feature attenuation at the brightness and chromaticity levels. Figure 3 , which is a quantitative difference comparison diagram of the histogram distribution of a clear remote sensing image and a haze remote sensing image in YUV space provided by an embodiment of the present invention.

[0036] Depend on Figure 2 , 3 It can be seen that compared with traditional RGB space processing, YUV space can effectively separate the brightness information (Y channel) and chromaticity information (U and V channels) of the image, which makes it possible to more accurately describe and correct degradation features. For example, the impact of haze on the brightness of the image is mainly reflected in the Y channel, while color distortion is mainly concentrated in the U and V channels. Therefore, by optimizing each channel in a targeted manner, the clarity and true color of the image can be restored more efficiently. Research on remote sensing image dehazing methods guided by YUV spatial features will not only help improve image quality and information extraction capabilities, but can also be widely used in environmental monitoring, urban planning, ecological protection and other fields, and promote the widespread implementation and innovative development of remote sensing technology in a variety of practical applications.

[0037] Therefore, before dehazing the haze image, it needs to be converted into a YUV image. According to the recommended standard (ITU-R BT.601-7) formulated by the International Telecommunication Union Radiocommunication Department, under full range conditions, the red (R), green (G), and blue (B) components in the RGB color model can be mapped to the brightness (Y) and chrominance signals (UV) in the YUV color model through linear transformation, thereby realizing the separation and re-expression of the image color information. The specific conversion formula is as follows:

[0038] The above space conversion process can be further abstracted into a matrix form to more intuitively represent the linear mapping relationship between RGB space and YUV space: .

[0039] After the conversion is completed, the three channels of the obtained YUV image can be disassembled to obtain the brightness component (Y channel component) and the chrominance components (U, V channel components).

[0040] For step S2, specifically, before performing defogging processing on the brightness component, it also includes: calculating the ambient light value and scene transmittance corresponding to the haze image, and the specific calculation process includes: Mapping the haze image to a grayscale space to obtain a grayscale image; From the grayscale image dark channel brightness b ‰ pixels, determine the pixel value of the pixel with the largest corresponding brightness value as the ambient light value; wherein, b is the preset ratio value; The scene transmittance is calculated according to the ambient light value and a preset atmospheric scattering model; wherein the calculation formula of the scene transmittance is specifically as follows:

[0041] in, is the scene transmittance, is the ambient light value, is the preset correction factor, is the preset atmospheric scattering model, To preset a clear image, The pixel point in the grayscale image A preset square area centered on the

[0042] In a preferred embodiment, for any image J(x) , and its corresponding dark channel characteristics can be described by the following formula:

[0043] in, Represents the pixels in the image x The local neighborhood window centered on Adaptively calculated based on brightness and regional characteristics:

[0044] in, μ and σ are the brightness mean and standard deviation, respectively, to ensure that the contribution of the highlight area to the dark channel image is weakened.

[0045] Based on the above formula, the improved dark channel prior algorithm calculation process can be constructed: 1) Under the assumption that the ambient light value A is a fixed constant, the atmospheric scattering model is mathematically derived, and its key characteristics are extracted by applying minimization processing on the left and right ends of the model. This process eliminates the interference of local fluctuations and highlights the core features in the scattering model, so that it can more accurately describe the brightness and color characteristics of the degraded image. The corresponding expression is as follows:

[0046] Transforming the above formula into:

[0047] Combined with the dark channel prior theory, we know that under the minimized state, The value of tends to 0, so the scene transmittance can be calculated based on the above equation:

[0048] in, w0 is a visual perception correction factor, which is set to 0.85 in the preferred embodiment.

[0049] 2) In the dark channel image, the brightest 1‰ pixels are selected according to the brightness sorting as candidate areas for estimating the global ambient light value A. Then, by analyzing the positions corresponding to these candidate pixels in the original image, the pixel size with the highest brightness value is selected as the specific value of A.

[0050] The improved dark channel prior algorithm proposed in the embodiment of the present invention combines the advantages of the traditional dark channel model with the limitations in practical applications, and introduces an adaptive adjustment mechanism to improve the accuracy of the algorithm in estimating feature parameters. Different from the original dark channel prior theory, the improved dark channel prior can effectively reduce the impact of the dark channel hypothesis on the abnormal brightness area by dynamically allocating weights to the abnormal brightness area, thereby avoiding excessive defogging or artifact generation.

[0051] For step S2, specifically, performing defogging on the brightness component specifically includes: According to the ambient light value and the scene transmittance, the brightness component is defogged to obtain a clear brightness component; wherein the specific calculation formula of the defogging process is:

[0052] in, is the clear brightness component, is the brightness component, is the preset stability constant.

[0053] Furthermore, the enhancing process of the chrominance component specifically includes: According to the ambient light value and the scene transmittance, the chromaticity component is enhanced to obtain a clear U channel component and a clear V channel component; wherein the specific calculation formula of the enhancement processing is:

[0054]

[0055] in, , are the clear U channel component and the clear V channel component respectively, , are respectively the U channel component and the V channel component in the chrominance component, is the preset stability constant.

[0056] In a preferred embodiment, by analyzing the degradation law of haze in different channels and combining the scattering and absorption characteristics in the physical model, a correction mechanism for brightness restoration and adaptive restoration calculation for chromaticity restoration are constructed:

[0057] in, It represents a fixed constant for maintaining calculation stability, which is set to 0.001 in the preferred embodiment.

[0058] Reference Figure 4 , is a schematic diagram of the visual representation of each channel in the YUV color space corresponding to a processed remote sensing image provided by an embodiment of the present invention, wherein (a) is a remote sensing haze image; (b), (c), and (d) correspond to the Y, U, and V spatial visualization features of the haze image, respectively; (e), (f), and (g) are the Y, U, and V spatial visualization features corresponding to the clear image restored by the method proposed by the present invention, respectively. Figure 4The brightness information of the image and the changes in the chromaticity components are presented in the form of decomposition of the Y, U, and V channels, which can more intuitively reflect the dehazing effect and visual enhancement effect of the improved algorithm on different color components, and verify the comprehensive performance of the algorithm under different channels.

[0059] For step S3, specifically, the restoration of the defogging brightness component and the enhanced chrominance component to obtain the defogging image specifically includes: According to the defogging brightness component and the enhanced chrominance component, a defogging YUV image is obtained by dimension stitching; The defogging YUV image is mapped from the YUV space to the RGB space to obtain the defogging image; wherein the mapping relationship from the YUV space to the RGB space is specifically: .

[0060] In a preferred embodiment, after the brightness component and the chrominance component are restored respectively, the restored YUV channel features can be dimensionally spliced, and the final clear dehazed image can be output through a matrix inverse mapping process from the YUV space to the RGB space.

[0061] In order to verify the effectiveness of the remote sensing image defogging method based on YUV spatial feature guidance described in the embodiment of the present invention, this preferred embodiment conducts experiments on the remote sensing haze image dataset DHID. The experimental hardware platform is Intel(R) Core(TM) i5-9400 CPU@2.90GHz, 16GB memory. The software platform is Win10 operating system, and the algorithm runs on Matlab2021a.

[0062] Reference Figure 5 , is a schematic diagram of the defogging processing result of a remote sensing image defogging method based on YUV spatial feature guidance provided by an embodiment of the present invention, wherein the first row of the image shows the original remote sensing haze image, and the last row shows the image output result obtained after adopting the defogging method proposed by the present invention. By comparison, it can be seen that the defogging method provided by the embodiment of the present invention effectively improves the visual quality of the remote sensing haze image. During the processing, the differential characteristics of the remote sensing image in brightness and chromaticity are comprehensively considered, and they are optimized in a targeted manner. While enhancing the perceived quality of the image and improving the visual clarity, the method described in the embodiment of the present invention also shows a strong generalization ability, which can adapt to the remote sensing image defogging requirements under different haze environments. These advantages not only improve the clarity and visualization effect of the image, but also provide effective support for the practical application of remote sensing image processing, especially in the fields of environmental monitoring and resource investigation, and have broad application prospects.

[0063] Reference Figure 6 , is a schematic structural diagram of a remote sensing image defogging device based on YUV spatial feature guidance provided by another embodiment of the present invention, comprising: a disassembly module 101, a processing module 102 and a restoration module 103; The decomposition module 101 is used to decompose the haze image to obtain the brightness component and the chromaticity component corresponding to the haze image; The processing module 102 is used to perform a defogging process on the brightness component and an enhancement process on the chrominance component; The restoration module 103 is used to restore the defogged brightness component and the enhanced chrominance component to obtain a defogged image.

[0064] Furthermore, the decomposition module 101 is used to decompose the haze image to obtain the brightness component and the chrominance component corresponding to the haze image, which specifically includes: The haze image is mapped from the RGB space to the YUV space to obtain a YUV haze image; wherein the mapping relationship from the RGB space to the YUV space is specifically: ; The YUV haze image is dimensionally decomposed to obtain the brightness component and the chrominance component; wherein the brightness component is the Y channel component corresponding to the YUV haze image, and the chrominance component is the U channel component and the V channel component corresponding to the YUV haze image.

[0065] Furthermore, before the processing module 102 performs defogging on the brightness component, it further includes: calculating the ambient light value and scene transmittance corresponding to the haze image, and the specific calculation process includes: Mapping the haze image to a grayscale space to obtain a grayscale image; From the grayscale image dark channel brightness b ‰ pixels, determine the pixel value of the pixel with the largest corresponding brightness value as the ambient light value; wherein, b is the preset ratio value; The scene transmittance is calculated according to the ambient light value and a preset atmospheric scattering model; wherein the calculation formula of the scene transmittance is specifically as follows:

[0066] in, is the scene transmittance, is the ambient light value, is the preset correction factor, is the preset atmospheric scattering model, The pixel point in the grayscale image A preset square area centered on the

[0067] Furthermore, the restoration module 103 is used to stitch and restore the defogged brightness component and the enhanced chrominance component to obtain a defogged image, which specifically includes: According to the defogging brightness component and the enhanced chrominance component, a defogging YUV image is obtained by dimension stitching; The defogging YUV image is mapped from the YUV space to the RGB space to obtain the defogging image; wherein the mapping relationship from the YUV space to the RGB space is specifically: .

[0068] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the embodiments of the present invention. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the embodiments here. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the claims of the present invention.

Claims

1. A remote sensing image defogging method based on YUV spatial feature guidance, characterized in that: The steps include: Decompose the haze image to obtain the brightness component and chromaticity component corresponding to the haze image; Performing a defogging process on the brightness component and an enhancement process on the chrominance component; The defogging brightness component and the enhanced chrominance component are stitched and restored to obtain a defogging image.

2. The remote sensing image defogging method based on YUV spatial feature guidance as claimed in claim 1, characterized in that: The decomposing of the haze image to obtain the brightness component and the chrominance component corresponding to the haze image specifically includes: The haze image is mapped from the RGB space to the YUV space to obtain a YUV haze image; wherein the mapping relationship from the RGB space to the YUV space is specifically: ； The YUV haze image is dimensionally decomposed to obtain the brightness component and the chrominance component; wherein the brightness component is the Y channel component corresponding to the YUV haze image, and the chrominance component is the U channel component and the V channel component corresponding to the YUV haze image.

3. The remote sensing image defogging method based on YUV spatial feature guidance as claimed in claim 1, characterized in that: Before performing defogging processing on the brightness component, the method further includes: calculating the ambient light value and scene transmittance corresponding to the haze image. The specific calculation process includes: Mapping the haze image to a grayscale space to obtain a grayscale image; From the grayscale image dark channel brightness b ‰ pixels, determine the pixel value of the pixel with the largest corresponding brightness value as the ambient light value; wherein, b is the preset ratio value; The scene transmittance is calculated according to the ambient light value and a preset atmospheric scattering model; wherein the calculation formula of the scene transmittance is specifically as follows: in, is the scene transmittance, is the ambient light value, is the preset correction factor, is the preset atmospheric scattering model, The pixel point in the grayscale image A preset square area centered on the 4. The remote sensing image defogging method based on YUV spatial feature guidance as claimed in claim 3, characterized in that: The defogging process for the brightness component specifically includes: According to the ambient light value and the scene transmittance, the brightness component is defogged to obtain a clear brightness component; wherein the specific calculation formula of the defogging process is: in, is the clear brightness component, is the brightness component, is the preset stability constant.

5. The remote sensing image defogging method based on YUV spatial feature guidance as claimed in claim 3, characterized in that: The enhancing process of the chrominance component specifically includes: According to the ambient light value and the scene transmittance, the chromaticity component is enhanced to obtain a clear U channel component and a clear V channel component; wherein the specific calculation formula of the enhancement processing is: in, , are the clear U channel component and the clear V channel component respectively, , are respectively the U channel component and the V channel component in the chrominance component, is the preset stability constant.

6. The remote sensing image defogging method based on YUV spatial feature guidance as claimed in claim 1, characterized in that: The step of restoring the defogging brightness component and the enhanced chrominance component to obtain a defogging image specifically includes: According to the defogging brightness component and the enhanced chrominance component, a defogging YUV image is obtained by dimension stitching; The defogging YUV image is mapped from the YUV space to the RGB space to obtain the defogging image; wherein the mapping relationship from the YUV space to the RGB space is specifically: 。 7. A remote sensing image defogging device based on YUV spatial feature guidance, characterized in that: include: Disassembly module, processing module and recovery module; The disassembly module is used to disassemble the haze image to obtain the brightness component and chromaticity component corresponding to the haze image; The processing module is used to perform a defogging process on the brightness component and an enhancement process on the chrominance component; The restoration module is used to restore the defogging brightness component and the enhanced chrominance component to obtain a defogging image.

8. The remote sensing image defogging device based on YUV spatial feature guidance as claimed in claim 7, characterized in that: The decomposition module is used to decompose the haze image to obtain the brightness component and chrominance component corresponding to the haze image, specifically including: The haze image is mapped from the RGB space to the YUV space to obtain a YUV haze image; wherein the mapping relationship from the RGB space to the YUV space is specifically: ； The YUV haze image is dimensionally decomposed to obtain the brightness component and the chrominance component; wherein the brightness component is the Y channel component corresponding to the YUV haze image, and the chrominance component is the U channel component and the V channel component corresponding to the YUV haze image.

9. The remote sensing image defogging device based on YUV spatial feature guidance as claimed in claim 7, characterized in that: Before the processing module performs defogging on the brightness component, the method further includes: calculating the ambient light value and scene transmittance corresponding to the haze image. The specific calculation process includes: Mapping the haze image to a grayscale space to obtain a grayscale image; From the grayscale image dark channel brightness b ‰ pixels, determine the pixel value of the pixel with the largest corresponding brightness value as the ambient light value; wherein, b is the preset ratio value; The scene transmittance is calculated according to the ambient light value and a preset atmospheric scattering model; wherein the calculation formula of the scene transmittance is specifically as follows: in, is the scene transmittance, is the ambient light value, is the preset correction factor, is the preset atmospheric scattering model, The pixel point in the grayscale image A preset square area centered on the 10. The remote sensing image defogging device based on YUV spatial feature guidance according to claim 7, characterized in that: The restoration module is used to stitch and restore the defogged brightness component and the enhanced chrominance component to obtain a defogged image, specifically comprising: According to the defogging brightness component and the enhanced chrominance component, a defogging YUV image is obtained by dimension stitching; The defogging YUV image is mapped from the YUV space to the RGB space to obtain the defogging image; wherein the mapping relationship from the YUV space to the RGB space is specifically: 。

Images