Underwater image enhancement method, system, equipment and medium

Abstract

The invention relates to an underwater image enhancement method, system and device and a medium, and is applied to image processing in water ecology monitoring, and the method comprises the steps: carrying out the multi-scale decomposition of a source image of an underwater environment, and obtaining a plurality of to-be-processed images with different resolutions; respectively carrying out detail enhancement on the source image and the plurality of to-be-processed images to obtain a plurality of detail images; carrying out fusion reconstruction on the plurality of detail images to obtain a preliminary enhanced image; and performing color correction on the preliminary enhanced image to obtain a target enhanced image. The problems that in the prior art, processing is generally limited to a single problem, the problems of water scattering and color distortion cannot be solved at the same time, adaptive adjustment cannot be carried out according to specific features of an underwater image, the image enhancement effect is not stable, and consequently the definition of the processed image is low, and the color deviation is large are solved.

Description

Background Art

[0002] In order to continuously improve the ecological protection and coordination pattern of the Yellow River Basin and build a strong national ecological security barrier. Water ecological monitoring has become an important focus of the current industry. Among them, macro-water ecological monitoring often involves underwater image acquisition, analysis and processing. Specifically, in underwater environments, the images captured often have problems such as low contrast, color distortion and blurred details. This is because when light propagates in water, light of various wavelengths is absorbed at different rates, among which red and orange light decay the fastest, while blue and green light can propagate farther. This uneven attenuation of the spectrum leads to color distortion in underwater images, which is usually manifested as the loss of red channel information and the overall image appears blue or green. In addition, as the water depth increases, the brightness of the light gradually decreases, resulting in reduced image contrast, blurred edges of objects, and lost details. Suspended particles in the water body can also cause light to scatter when passing through the water body, resulting in blurred objects in the image and uneven light. These problems are particularly evident in turbid water bodies, deep seas and lakes in the Yellow River Basin, bringing great challenges to tasks such as underwater monitoring, aquatic biological research, and fish population analysis. To solve these problems, traditional underwater image processing methods include the use of image denoising, contrast enhancement, white balance adjustment and histogram equalization techniques.

[0003] However, these existing technologies are usually limited to dealing with a single problem, unable to solve the problems of water scattering and color distortion at the same time, and unable to make adaptive adjustments based on the specific characteristics of underwater images. The image enhancement effect is unstable, resulting in low clarity and large color deviation of the processed image. For example, the color correction method based on physical models can restore colors to a certain extent, but has limited effect on improving color details and clarity. Summary of the invention

[0004] In order to overcome the problems that the prior art is usually limited to processing a single problem, cannot solve the problems of water scattering and color distortion at the same time, and cannot perform adaptive adjustments according to the specific characteristics of underwater images, the image enhancement effect is unstable, resulting in low clarity and large color deviation of the processed image, the present application provides an underwater image enhancement method, system, device and medium.

[0005] In a first aspect, in order to solve the above technical problems, the present application provides an underwater image enhancement method, comprising:

[0006] Perform multi-scale decomposition on the source image of the underwater environment to obtain multiple images to be processed with different resolutions;

[0007] Performing detail enhancement on a source image and a plurality of images to be processed respectively to obtain a plurality of detail images;

[0008] Multiple detail images are fused and reconstructed to obtain a preliminary enhanced image;

[0009] Perform color correction on the preliminary enhanced image to obtain the target enhanced image.

[0010] In a second aspect, the present application also provides an underwater image enhancement system, comprising:

[0011] A multi-scale decomposition module is used to perform multi-scale decomposition on the source image of the underwater environment to obtain multiple images to be processed with different resolutions;

[0012] A detail enhancement module, used for performing detail enhancement on a source image and a plurality of images to be processed respectively, to obtain a plurality of detail images;

[0013] A fusion module is used to fuse and reconstruct multiple detail images to obtain a preliminary enhanced image;

[0014] The color correction module is used to perform color correction on the preliminary enhanced image to obtain a target enhanced image.

[0015] In a third aspect, the present application also provides a computing device, including a memory, a processor, and a program stored in the memory and running on the processor, and when the processor executes the program, the steps of an underwater image enhancement method as described above are implemented.

[0016] In a fourth aspect, the present application also provides a computer-readable storage medium, in which instructions are stored. When the instructions are executed on a terminal device, the terminal device executes the steps of a method for enhancing an underwater image.

[0017] The beneficial effect of the present application is: by performing multi-scale decomposition processing on the source image of the underwater environment, and then performing detail enhancement, the detail performance of the source image is improved, so that the detail information that can meet the visual recognition requirements can be retained in the obtained multiple detail images. Since the more detail information an image contains, the clearer the image is, the multiple detail images are fused and reconstructed into a preliminary enhanced image, that is, the detail information in each detail image is fused into the preliminary enhanced image, so that the clarity of the preliminary enhanced image can be improved, so that the preliminary enhanced image can meet the clarity requirements of visual recognition. Finally, the preliminary enhanced image is color corrected, which can reduce the color deviation between the obtained target enhanced image and the real image. In this way, the method of the present application can automatically perform detail enhancement and color correction adjustment based on a source image to stabilize the detail information and color information of the enhanced image, which can not only solve the problems of water scattering and color distortion of the source image of the underwater environment at the same time, but also improve the clarity of the processed target enhanced image, reduce its color deviation, and thus improve the accuracy of the target enhanced image. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 A schematic flow chart of an underwater image enhancement method shown as an exemplary embodiment of the present application;

[0019] Figure 2 This is a flowchart of an underwater image enhancement method provided in an exemplary embodiment of the present application;

[0020] Figure 3 A comparison diagram of a target enhanced image and a source image obtained by applying a provided underwater image enhancement method in an exemplary embodiment of the present application;

[0021] Figure 4 This is a flowchart of a detail enhancement step in a method for enhancing an underwater image provided in an exemplary embodiment of the present application;

[0022] Figure 5 The present invention is a schematic structural diagram of an underwater image enhancement system according to an exemplary embodiment of the present application. DETAILED DESCRIPTION

[0023] The following examples are provided to further explain and supplement the present application and do not constitute any limitation to the present application.

[0024] The following describes an underwater image enhancement method, system, device and medium according to an embodiment of the present application in conjunction with the accompanying drawings.

[0025] An underwater image enhancement method of an embodiment of the present application is applied to a terminal device. In the present application scheme, the terminal device is used as the execution subject to illustrate the present application scheme. The terminal device is used to execute the steps of an underwater image enhancement method.

[0026] See also Figure 1 , Figure 1 An underwater image enhancement method is shown as an exemplary embodiment of the present application. Figure 1 As shown, the present application provides a method for enhancing an underwater image, comprising:

[0027] S11, performing multi-scale decomposition on the source image of the underwater environment to obtain a plurality of images to be processed with different resolutions;

[0028] S12, performing detail enhancement on the source image and the multiple images to be processed respectively to obtain multiple detail images;

[0029] S13, fusing and reconstructing multiple detail images to obtain a preliminary enhanced image;

[0030] S14, performing color correction on the preliminary enhanced image to obtain a target enhanced image.

[0031] The method for enhancing an underwater image of the present embodiment performs multi-scale decomposition processing on the source image of the underwater environment, and then performs detail enhancement to improve the detail performance of the source image, so that the detail information that can meet the visual recognition requirements can be retained in the obtained multiple detail images. Since the more detail information an image contains, the clearer the image is, the multiple detail images are fused and reconstructed into a preliminary enhanced image, that is, the detail information in each detail image is fused into the preliminary enhanced image, so that the clarity of the preliminary enhanced image can be improved, so that the preliminary enhanced image can meet the clarity requirements of visual recognition. Finally, the preliminary enhanced image is color corrected to reduce the color deviation between the obtained target enhanced image and the real image. In this way, the method of the present embodiment can automatically perform detail enhancement and color correction adjustment based on a source image to stably enhance the detail information and color information of the image, which can not only solve the problems of water scattering and color distortion of the source image of the underwater environment at the same time, but also improve the clarity of the processed target enhanced image, reduce its color deviation, and thus improve the accuracy of the target enhanced image.

[0032] In this embodiment, multi-scale decomposition refers to a processing method for multi-scale decomposition of a source image using Gaussian pyramid technology. The specific processing method is: by downsampling the source image layer by layer, a plurality of images to be processed with different resolutions are generated, thereby capturing the global and local features of the source image at different scales, so as to retain a sufficient number of features in the source image.

[0033] Fusion reconstruction refers to the use of Laplace pyramid fusion technology to reconstruct the image information in multiple detail images after detail enhancement of the source image and the processed images of different resolutions at various scales. The reconstructed preliminary enhanced image combines the detail information of the source image at different scales, especially the subtle features of various underwater animals, plants and other substances. It can enhance the edges and details of various underwater animals, plants and other substances while maintaining the overall structure of the source image, so that the preliminary enhanced image can meet the clarity and contrast requirements of visual recognition.

[0034] See also Figure 2 , Figure 2 This is a flow chart of an underwater image enhancement method provided by the present application in an exemplary embodiment, such as Figure 2As shown, the source image of the underwater environment is first acquired, and then the source image and the multiple images to be processed obtained by multi-scale decomposition are enhanced in detail to improve the clarity and contrast of the source image, and then the preliminary enhanced image obtained by fusing and reconstructing the multiple detail images after detail enhancement is color corrected to restore the color balance of the source image, thereby reducing the color deviation between the enhanced target enhanced image of the underwater environment and the real image. In this way, the clarity of the enhanced target enhanced image obtained by processing the source image of the underwater environment by the image enhancement method of this embodiment is improved and the color deviation is reduced, thereby improving the visual quality of the target enhanced image. In this embodiment, the comparison diagram between the enhanced target enhanced image and the source image is shown in FIG. Figure 3 shown.

[0035] Optionally, the plurality of images to be processed include a first resolution image and a second resolution image;

[0036] The source image and the multiple images to be processed are respectively enhanced in detail to obtain multiple detail images, including:

[0037] Based on the image resolution, obtaining detail enhancement strategies corresponding to the source image, the first resolution image, and the second resolution image respectively;

[0038] For the source image, a detail enhancement strategy matching the source image is executed to obtain a first image;

[0039] For the first resolution image, executing a detail enhancement strategy matching the first resolution image to obtain a second image;

[0040] For the second resolution image, executing a detail enhancement strategy matching the second resolution image to obtain a third image;

[0041] A plurality of detail images are obtained based on the first image, the second image and the third image.

[0042] In this embodiment, detail enhancement strategies corresponding to the image resolution are set for the source image, the first resolution image, and the second resolution image, respectively, and the detail enhancement strategies matching the respective images are executed accordingly to obtain the first image, the second image, and the third image, which are combined into a plurality of detail images. In this way, by performing detail enhancement processing on the source image at different scales, the detail information of the source image at different scales can be fully utilized, thereby improving the detail performance of the source image, so that the detail information that can meet the visual recognition requirements can be retained in the obtained plurality of detail images, thereby improving the detail clarity of the plurality of detail images while avoiding image distortion caused by excessive enhancement, facilitating the subsequent improvement of the clarity and overall contrast of the preliminary enhanced image obtained by fusing and reconstructing the plurality of detail images, so that the preliminary enhanced image can meet the clarity requirements of visual recognition.

[0043] Optionally, executing a detail enhancement strategy matching the source image to obtain a first image includes:

[0044] Perform bilateral filtering on the source image to obtain a filtered image;

[0045] Performing discrete wavelet decomposition on the filtered image to obtain a first low-frequency component and a first high-frequency component;

[0046] The first low-frequency component and the first high-frequency component are reconstructed using discrete wavelet to obtain a first image.

[0047] In this embodiment, bilateral filtering is a nonlinear filtering method that can effectively remove noise in an image while retaining edge texture details. The low-frequency component is a parameter that maintains the consistency of the overall illumination and main structure of the image, and the high-frequency component is the two parameters of detail information and edge information obtained by extracting and analyzing the multi-scale detail information of the image and fusing them. This embodiment removes the noise in the source image by performing bilateral filtering on the source image, thereby reducing the interference of suspended particles in the source image, while retaining the edge detail information and texture detail information of the source image, so that the effective information in the obtained filtered image can be retained. Then, the filtered image is decomposed by discrete wavelet to obtain a first low-frequency component and a first high-frequency component; the first low-frequency component and the first high-frequency component are reconstructed by discrete wavelet to obtain a first image, so that the first image can contain the overall illumination information, main structure information, detail information and edge information in the filtered image, thereby improving the detail expression ability of the first image.

[0048] Discrete wavelet decomposition and discrete wavelet reconstruction are two forms of discrete wavelet transform (DWT). Discrete wavelet transform (DWT) mainly changes the scale of the image, and can more deeply extract various detailed features of images at different scales: overall illumination features, main structural features, detail features, and edge features. Discrete wavelet decomposition technology is used to provide a deeper analysis of the multi-level feature information of the image, thereby further enhancing the detail expression of various particles in the image while maintaining the overall structure of the image.

[0049] Optionally, executing a detail enhancement strategy matching the first resolution image to obtain a second image includes:

[0050] Performing stable wavelet decomposition on the first resolution image to obtain a second low-frequency component;

[0051] The second low-frequency component and the first high-frequency component are reconstructed using stable wavelet to obtain a second image.

[0052] In this embodiment, the second low-frequency component obtained by performing stable wavelet decomposition on the first resolution image and the first high-frequency component obtained by performing discrete wavelet decomposition on the filtered image are subjected to stable wavelet reconstruction, and the second low-frequency component and the first high-frequency component can be aligned and fused into a second image, thereby improving the fusion degree of the obtained second image. At the same time, the obtained second image not only retains the detail information and edge information in the source image, but also retains the detail information and edge information in the second resolution image, so as to realize the fusion of detail information and edge information of two different scales, so that the second image can have detail information and edge information of two scales, thereby improving the detail expression ability of the second image.

[0053] Stable wavelet decomposition and stable wavelet reconstruction are two forms of stable wavelet transform (SWT). Stable wavelet decomposition is used to preserve the edges and details of the image without changing the image size, ensuring the alignment between images of different scales, so as to improve the fusion degree when reconstructing multi-scale images.

[0054] Optionally, executing a detail enhancement strategy matching the second resolution image to obtain a third image includes:

[0055] Performing stable wavelet decomposition on the second resolution image to obtain a third low-frequency component;

[0056] Performing discrete wavelet decomposition on the second image to obtain a second high-frequency component;

[0057] The second high-frequency component and the third low-frequency component are reconstructed using stable wavelet to obtain a third image.

[0058] In this embodiment, the third low-frequency component obtained by performing stable wavelet decomposition on the second resolution image and the second high-frequency component obtained by performing discrete wavelet decomposition on the second image are reconstructed by stable wavelet, and the third low-frequency component and the second high-frequency component can be aligned and fused into a third image, thereby improving the fusion degree of the obtained third image. At the same time, the obtained third image not only retains the detail information and edge information in the second resolution image, but also retains the detail information and edge information in the second image obtained by fusion of image components of two different scales, so as to realize the fusion of detail information and edge information of three different scales, so that the third image can have detail information and edge information of three scales, thereby improving the detail expression ability of the third image.

[0059] See also Figure 4 , Figure 4 This is a schematic diagram of a process of detail enhancement in a method for enhancing an underwater image provided in an exemplary embodiment of the present application, such as Figure 4As shown, the resolution of the source image A of the underwater environment is 1024×1024. The source image A is decomposed at multiple scales, and Gaussian down-sampling is performed on the lower two layers in turn to obtain a first resolution image G1 with a resolution of 512×512 and a second resolution image G2 with a resolution of 256×256.

[0060] Then, bilateral filtering is performed on the source image A to obtain a filtered image A1 with a resolution of 1024×1024. Discrete wavelet decomposition is performed on the filtered image A1 to obtain a first low-frequency component and a first high-frequency component with a resolution of 512×512. The first low-frequency component and the first high-frequency component are reconstructed by discrete wavelet to obtain a first image E1 with a resolution of 1024×1024. Stable wavelet decomposition is performed on the first-resolution image G1 to obtain a second low-frequency component and a third high-frequency component with a resolution of 512×512. Stable wavelet reconstruction is performed on the second low-frequency component and the first high-frequency component to obtain a second image E2 with a resolution of 512×512. Stable wavelet decomposition is performed on the second-resolution image G1 to obtain a third low-frequency component and a fourth high-frequency component with a resolution of 256×256. Discrete wavelet decomposition is performed on the second image to obtain a fourth low-frequency component and a second high-frequency component with a resolution of 256×256. Stable wavelet reconstruction is performed on the second high-frequency component and the third low-frequency component to obtain a third image E3 with a resolution of 256×256.

[0061] Finally, the Laplacian pyramid fusion technology is used to fuse and reconstruct the first image E1 with a resolution of 1024×1024, the second image E2 with a resolution of 512×512, and the third image E3 with a resolution of 256×256 to obtain a preliminary enhanced image E.

[0062] Optionally, color correction is performed on the preliminary enhanced image to obtain a target enhanced image, including:

[0063] For each pixel in the preliminary enhanced image, obtain the red channel intensity, green channel intensity and blue channel intensity of the pixel;

[0064] Performing color correction on the pixel points based on the red channel intensity, the green channel intensity and the blue channel intensity to obtain the target enhanced pixel points of the pixel points;

[0065] An image composed of multiple target enhanced pixel points corresponding to the preliminary enhanced image is used as the target enhanced image.

[0066] In this embodiment, due to the absorption and scattering of light in an underwater environment, light of different wavelengths is absorbed at different speeds, especially red light decays the fastest, while blue light and green light can penetrate deeper water layers. Therefore, the source image of the underwater environment captured and collected usually presents a bluish or greenish tone and lacks a natural balance of colors. The preliminary enhanced image obtained by detail enhancement and fusion reconstruction of the source image only meets the clarity requirements of visual recognition, and the preliminary enhanced image still presents a bluish or greenish tone, and has a large color deviation from the real image. Therefore, it is necessary to perform targeted color correction on each pixel point according to the red channel intensity, green channel intensity and blue channel intensity of each pixel point in the preliminary enhanced image, so as to flexibly adjust the compensation intensity of the red channel in the preliminary enhanced image of the underwater environment, and avoid color distortion caused by overcorrection. In this way, the color distortion phenomenon caused by underwater light attenuation and scattering can be improved, the natural colors of the source image and the preliminary enhanced image can be restored, the color deviation between the target enhanced image obtained by processing and the real image can be reduced, and the overall balance and authenticity of the color in the target enhanced image obtained can be ensured, thereby providing a better visual effect.

[0067] Optionally, color correction is performed on the pixel based on the red channel intensity, the green channel intensity, and the blue channel intensity to obtain a target enhanced pixel of the pixel, including:

[0068] Obtain the average pixel value of the red channel, the average pixel value of the green channel and the average pixel value of the blue channel of the preliminary enhanced image;

[0069] Based on the red channel intensity, the green channel intensity, the blue channel intensity, the average red channel intensity, the average green channel intensity and the average blue channel intensity, the target red channel intensity of the pixel is obtained;

[0070] The pixel points are color corrected based on the target red channel intensity to obtain the target enhanced pixel points.

[0071] In this embodiment, for each pixel point, the target red channel intensity of the pixel point is obtained based on the corresponding red channel intensity, green channel intensity, blue channel intensity, red channel intensity average, green channel intensity average and blue channel intensity average, and the pixel point is color corrected based on the target red channel intensity, which can compensate for the red information lost in the preliminary enhanced image due to underwater light attenuation to restore the red detail information in the preliminary enhanced image, so that the target enhanced image composed of each target enhanced pixel point obtained after color correction has a smaller color deviation from the corresponding real image and is closer to the original natural state.

[0072] In this embodiment, the red channel intensity of each pixel is adjusted based on the necessary adjustment required because the red intensity of each pixel is affected by environmental factors. This is not just a simple "retention" of the red channel information, but a dynamic adjustment to make it closer to the color performance under natural light.

[0073] Optionally, the target red channel intensity is calculated as follows:

[0074]

[0075] Among them, E RN (x) represents the red channel after color correction, E R 、E G 、E B They represent the red channel, green channel and blue channel of the image respectively, and x is the pixel point of the corresponding channel. They represent the pixel averages of the red channel, the green channel, and the blue channel in the source image or the preliminary enhanced image, respectively. R (x), E G (x), E B (x) represents the red channel intensity, green channel intensity, and blue channel intensity of a pixel in the source image or the preliminary enhanced image, respectively.

[0076] In this embodiment, by calculating the difference between the average of the pixel average of the green channel and the pixel average of the blue channel and the pixel average of the red channel, the global deviation degree of the red channel intensity of each pixel in the source image in the underwater environment relative to the other channel intensities in the entire image can be reflected. Due to the rapid attenuation of red light in the underwater environment, the global deviation degree corresponding to the source image obtained by shooting is relatively large, which means that the red channel intensity of each pixel in the source image or the preliminary enhanced image needs to be adjusted to a large extent to restore the color balance, so as to achieve global channel difference compensation (color compensation) of the source image or the preliminary enhanced image.

[0077] At the same time, (1-E R (x)) and Both are local pixel adjustment factors, and they work together to adjust the intensity of the red channel of the pixel (target red channel intensity). (1-E R (x)) is a dynamic adjustment factor that takes into account that if a pixel's red channel is brighter, then it does not need as much compensation to avoid oversaturation; conversely, areas with darker red channels require more compensation. It is used to adjust the compensation intensity of the red channel intensity according to the average value of the green channel intensity and the blue channel intensity of a pixel point, ensuring the consistency of the compensation with the local color environment and improving the naturalness and accuracy of color correction.

[0078] In this way, the present embodiment flexibly adjusts the red channel of each pixel in the preliminary enhanced image according to its specific situation. This fine pixel-level adjustment ensures the overall consistency and naturalness of the image color. Compared with the traditional simple global color adjustment method, the color correction method of the present embodiment makes adjustments at the local pixel level to avoid oversaturation or distortion, and can more accurately handle the complex light and color conditions of underwater images, effectively preventing oversaturation or undersaturation of colors, improving the overall visual quality of the image, and making the underwater image closer to the color performance under normal lighting conditions on land. In addition, the color correction method of the present embodiment provides strong technical support for aquatic biological research, water quality monitoring, and underwater environmental detection, and has important practical significance.

[0079] The present invention discloses an underwater image enhancement method, which performs multi-scale decomposition and detail enhancement on the source image of the underwater environment through multi-scale decomposition, bilateral filtering, wavelet transform, and fusion reconstruction techniques, and accurately optimizes the extraction of local details and global structures of the image. Among them, multi-scale decomposition captures the features of different levels of the image, bilateral filtering retains the edge details of the image and effectively removes noise, and the combination of stable wavelet transform (SWT) and discrete wavelet transform (DWT) further enhances the detail expression of the image. In the image fusion and reconstruction process, by fusing the detail information of images of different scales, it is ensured that the local details are enhanced without losing the overall structure, thereby improving the clarity and contrast of the image. In this way, while maintaining the global structural stability of the source image, the local details can be significantly enhanced, so that the image can show a better visual effect in a complex underwater environment, thereby improving the clarity, contrast and detail expression of the source image. The initial enhanced image obtained after detail enhancement is color corrected for the missing red channel information, and the red channel intensity of each pixel in the initial enhanced image is dynamically adjusted to compensate for the color deviation of the initial enhanced image caused by underwater water scattering and light attenuation, ensuring the naturalness and realism of the color of the target enhanced image obtained, solving the color deviation problem caused by water scattering and light attenuation, improving image quality, and making the color of the final target enhanced image more natural and realistic, thereby significantly improving the overall clarity, detail expression and visual quality of the source image of the underwater environment. The method of the present application is suitable for multiple application scenarios such as underwater monitoring, marine ecological research and fish behavior analysis.

[0080] The advantages of an underwater image enhancement method of the present application are:

[0081] Detail enhancement: Through the combination of multi-scale decomposition and wavelet transform, the local details and global structure of the image are effectively preserved and enhanced, especially in complex underwater environments, and the subtle features of underwater organisms, animals and plants can be clearly presented.

[0082] Noise Removal: Bilateral filtering technology is used to reduce noise while retaining important edge information, avoiding the loss of details caused by traditional denoising methods.

[0083] Color correction: It can dynamically adjust the red channel intensity of each pixel in the image to compensate for the color distortion caused by underwater light, making the image color more natural and enhancing the visual effect of the image.

[0084] Global and local optimization: The present invention can optimize both the overall structure and local details of the image, ensuring the uniformity and accuracy of the visual effects.

[0085] See also Figure 5 , Figure 5 An underwater image enhancement system is shown as an exemplary embodiment of the present application. Figure 5 As shown, the present application provides an underwater image enhancement system, comprising:

[0086] A multi-scale decomposition module is used to perform multi-scale decomposition on the source image of the underwater environment to obtain multiple images to be processed with different resolutions;

[0087] A detail enhancement module, used for performing detail enhancement on a source image and a plurality of images to be processed respectively, to obtain a plurality of detail images;

[0088] A fusion module is used to fuse and reconstruct multiple detail images to obtain a preliminary enhanced image;

[0089] The color correction module is used to perform color correction on the preliminary enhanced image to obtain a target enhanced image.

[0090] An underwater image enhancement system of this embodiment performs multi-scale decomposition processing on the source image of the underwater environment through a multi-scale decomposition module, and then uses a detail enhancement module to enhance the details, improve the detail performance of the source image, so that the obtained multiple detail images can retain detail information that can meet the visual recognition requirements. Since the more detail information an image contains, the clearer the image is, the fusion module is used to fuse and reconstruct multiple detail images into a preliminary enhanced image, that is, the detail information in each detail image is fused into the preliminary enhanced image, so that the clarity of the preliminary enhanced image can be improved, so that the preliminary enhanced image can meet the clarity requirements of visual recognition. Finally, the color correction module is used to perform color correction on the preliminary enhanced image, which can reduce the color deviation between the obtained target enhanced image and the real image. In this way, the system of this embodiment can automatically perform detail enhancement and color correction adjustment based on a source image to stably enhance the detail information and color information of the image, which can not only solve the problems of water scattering and color distortion of the source image of the underwater environment at the same time, but also improve the clarity of the processed target enhanced image, reduce its color deviation, and thus improve the accuracy of the target enhanced image.

[0091] Optionally, the multiple images to be processed include a first resolution image and a second resolution image; and the detail enhancement module is specifically used to:

[0092] Based on the image resolution, obtaining detail enhancement strategies corresponding to the source image, the first resolution image, and the second resolution image respectively;

[0093] For the source image, a detail enhancement strategy matching the source image is executed to obtain a first image;

[0094] For the first resolution image, executing a detail enhancement strategy matching the first resolution image to obtain a second image;

[0095] For the second resolution image, executing a detail enhancement strategy matching the second resolution image to obtain a third image;

[0096] A plurality of detail images are obtained based on the first image, the second image and the third image.

[0097] Optionally, the detail enhancement module is specifically used to:

[0098] Perform bilateral filtering on the source image to obtain a filtered image;

[0099] Performing discrete wavelet decomposition on the filtered image to obtain a first low-frequency component and a first high-frequency component;

[0100] The first low-frequency component and the first high-frequency component are reconstructed using discrete wavelet to obtain a first image.

[0101] Optionally, the detail enhancement module is specifically used to:

[0102] Performing stable wavelet decomposition on the first resolution image to obtain a second low-frequency component;

[0103] The second low-frequency component and the first high-frequency component are reconstructed using stable wavelet to obtain a second image.

[0104] Optionally, the detail enhancement module is specifically used to:

[0105] Performing stable wavelet decomposition on the second resolution image to obtain a third low-frequency component;

[0106] Performing discrete wavelet decomposition on the second image to obtain a second high-frequency component;

[0107] The second high-frequency component and the third low-frequency component are reconstructed using stable wavelet to obtain a third image.

[0108] Optionally, the color correction module is specifically used to:

[0109] For each pixel in the preliminary enhanced image, obtain the red channel intensity, green channel intensity and blue channel intensity of the pixel;

[0110] Performing color correction on the pixel points based on the red channel intensity, the green channel intensity and the blue channel intensity to obtain the target enhanced pixel points of the pixel points;

[0111] An image composed of multiple target enhanced pixel points corresponding to the preliminary enhanced image is used as the target enhanced image.

[0112] Optionally, the color correction module is specifically used to:

[0113] Obtain the average pixel value of the red channel, the average pixel value of the green channel and the average pixel value of the blue channel of the preliminary enhanced image;

[0114] Based on the red channel intensity, the green channel intensity, the blue channel intensity, the average red channel intensity, the average green channel intensity and the average blue channel intensity, the target red channel intensity of the pixel is obtained;

[0115] The pixel points are color corrected based on the target red channel intensity to obtain the target enhanced pixel points.

[0116] A computing device according to an embodiment of the present application includes a memory, a processor, and a program stored in the memory and running on the processor. When the processor executes the program, part or all of the steps of the above-mentioned underwater image enhancement method are implemented.

[0117] Among them, the computing device can be a computer, and correspondingly, its program is computer software. The above-mentioned parameters and steps in a computing device of the present application can refer to the parameters and steps in an embodiment of an underwater image enhancement method mentioned above, and will not be repeated here.

[0118] In an embodiment of the present application, a computer-readable storage medium is provided, in which instructions are stored. When the instructions are executed, the steps of the above-mentioned underwater image enhancement method are executed.

[0119] The computer-readable storage medium may be a transient computer-readable storage medium or a non-transitory computer-readable storage medium.

[0120] The technical solution of the embodiment of the present disclosure can be embodied in the form of a software product, which is stored in a storage medium and includes one or more instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method of the embodiment of the present disclosure. The aforementioned computer-readable storage medium may be a non-transitory computer-readable storage medium, including: a USB flash drive, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a disk or an optical disk, and other media that can store program codes, or a transient computer-readable storage medium.

[0121] The flowchart and block diagram in the accompanying drawings illustrate the possible architecture, functions and operations of the system, method and computer program product according to various embodiments of the present application. Wherein, each box in the flowchart or block diagram can represent a module, a program segment, or a part of the code, and the above-mentioned module, program segment, or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flowchart, and the combination of boxes in the block diagram or flowchart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0122] Those skilled in the art know that the present application can be implemented as a system, method or computer program product. Therefore, the present disclosure can be specifically implemented in the following forms, namely: it can be complete hardware, it can be complete software (including firmware, resident software, microcode, etc.), or it can be a combination of hardware and software, generally referred to as "module" or "system" herein. In addition, in some embodiments, the present application can also be implemented in the form of a computer program product in one or more computer-readable media, and the computer-readable medium contains computer-readable program code. Computer-readable storage media can be, for example, but not limited to - electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or devices, or any combination of the above.

[0123] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0124] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations on the present application. Ordinary technicians in this field can change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A method for enhancing underwater images, characterized in that: include: Perform multi-scale decomposition on the source image of the underwater environment to obtain multiple images to be processed with different resolutions; Performing detail enhancement on the source image and multiple images to be processed respectively to obtain multiple detail images; fusing and reconstructing the multiple detail images to obtain a preliminary enhanced image; Color correction is performed on the preliminary enhanced image to obtain a target enhanced image.

2. The method according to claim 1, characterized in that The multiple images to be processed include a first resolution image and a second resolution image; The step of performing detail enhancement on the source image and the plurality of images to be processed respectively to obtain a plurality of detail images comprises: Based on the image resolution, obtaining detail enhancement strategies corresponding to the source image, the first resolution image, and the second resolution image respectively; For the source image, executing a detail enhancement strategy matching the source image to obtain a first image; For the first resolution image, executing a detail enhancement strategy matching the first resolution image to obtain a second image; For the second resolution image, executing a detail enhancement strategy matching the second resolution image to obtain a third image; A plurality of detail images are obtained based on the first image, the second image and the third image.

3. The method according to claim 2, characterized in that The step of executing a detail enhancement strategy matching the source image to obtain a first image includes: Performing bilateral filtering on the source image to obtain a filtered image; Performing discrete wavelet decomposition on the filtered image to obtain a first low-frequency component and a first high-frequency component; Perform discrete wavelet reconstruction on the first low-frequency component and the first high-frequency component to obtain a first image.

4. The method according to claim 3, characterized in that The step of executing a detail enhancement strategy matching the first resolution image to obtain a second image includes: Performing stable wavelet decomposition on the first resolution image to obtain a second low-frequency component; Perform stable wavelet reconstruction on the second low-frequency component and the first high-frequency component to obtain a second image.

5. The method according to claim 4, characterized in that The step of executing a detail enhancement strategy matching the second resolution image to obtain a third image includes: performing stable wavelet decomposition on the second resolution image to obtain a third low-frequency component; performing discrete wavelet decomposition on the second image to obtain a second high-frequency component; Perform stable wavelet reconstruction on the second high-frequency component and the third low-frequency component to obtain a third image.

6. The method according to any one of claims 1 to 5, characterized in that: The color correction of the preliminary enhanced image to obtain a target enhanced image includes: For each pixel in the preliminary enhanced image, obtaining the red channel intensity, green channel intensity and blue channel intensity of the pixel; Performing color correction on the pixel point based on the red channel intensity, the green channel intensity, and the blue channel intensity to obtain a target enhanced pixel point of the pixel point; An image composed of a plurality of target enhanced pixel points corresponding to the preliminary enhanced image is used as a target enhanced image.

7. The method according to claim 6, characterized in that The color correction of the pixel point based on the red channel intensity, the green channel intensity, and the blue channel intensity to obtain the target enhanced pixel point of the pixel point includes: Obtaining an average value of red channel pixels, an average value of green channel pixels, and an average value of blue channel pixels of the preliminary enhanced image; Based on the red channel intensity, the green channel intensity, the blue channel intensity, the red channel intensity average, the green channel intensity average and the blue channel intensity average, a target red channel intensity of the pixel is obtained; The pixel point is color corrected based on the target red channel intensity to obtain a target enhanced pixel point.

8. An underwater image enhancement system, characterized in that: include: A multi-scale decomposition module is used to perform multi-scale decomposition on the source image of the underwater environment to obtain multiple images to be processed with different resolutions; A detail enhancement module, used for performing detail enhancement on the source image and a plurality of images to be processed respectively, to obtain a plurality of detail images; A fusion module, used for fusing and reconstructing the multiple detail images to obtain a preliminary enhanced image; The color correction module is used to perform color correction on the preliminary enhanced image to obtain a target enhanced image.

9. A computing device comprising a memory, a processor, and a program stored in the memory and running on the processor, characterized in that: When the processor executes the program, the steps of the underwater image enhancement method as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores instructions, and when the instructions are executed on a terminal device, the terminal device executes the steps of an underwater image enhancement method as described in any one of claims 1 to 7.

Images