Low-light image adaptive enhancement method, device, storage medium and electronic device

Abstract

The present application discloses a method, device, storage medium, and electronic device for adaptively enhancing low-light images, relating to the field of image processing technology. The method comprises: decomposing a low-light image to be enhanced using a discrete wavelet transform to obtain sub-band images, wherein the sub-band images include a first image representing low-frequency color features and multiple second images representing high-frequency detail features; enhancing the first image using an automatic color scale enhancement method to obtain a first enhanced image; enhancing each second image using a fast Fourier transform method to obtain a second enhanced image corresponding to each second image; reconstructing the first enhanced image and each second enhanced image using an inverse discrete wavelet transform reconstruction method to obtain an initial low-light enhanced image; and enhancing the initial low-light enhanced image using a structured weighted guided filtering method to obtain a target low-light enhanced image. The method of the present application can improve the quality of low-light image enhancement.

Description

Technical Field

[0001] The present invention relates to the field of image enhancement, and in particular to a method, device, storage medium and electronic device for adaptively enhancing a low-light image. Background Art

[0002] Image enhancement is one of the most important techniques in image processing, aiming to improve image quality for specific applications. Generally speaking, the fundamental principle of image enhancement is to improve the quality and visual interpretability of images, making them more suitable for specific applications and observers. However, due to unavoidable factors in the image acquisition process, images captured under natural conditions often suffer from various degradation issues due to factors such as noise and fog, resulting in blurry images, color distortion, and various image defects. The quality of these low-light images deteriorates significantly after image conversion, storage, and transmission, making them difficult to meet engineering or research needs.

[0003] Traditional image enhancement methods include spatial domain-based and frequency domain-based image enhancement methods. With the development of deep learning, deep learning-based image enhancement strategies have gradually become mainstream. In addition, there are also methods based on image fusion.

[0004] The above-mentioned image enhancement methods can enhance traditional images, but they are not very effective for visible light images in low-light environments. Therefore, how to enhance visible light images in low-light environments has become an urgent problem to be solved. Summary of the Invention

[0005] In view of this, the present invention provides a method, device, storage medium and electronic device for adaptively enhancing low-light images, the main purpose of which is to solve the current problem of low quality of visible light image enhancement in low-light environments.

[0006] To solve the above problems, the present application provides a method for adaptively enhancing low-light images, comprising:

[0007] Decomposing the low-light image to be enhanced by discrete wavelet transform to obtain sub-band images, wherein the sub-band images include a first image representing low-frequency color features and a plurality of second images representing high-frequency detail features;

[0008] Performing enhancement processing on the first image using an automatic color scale enhancement method to obtain a first enhanced image;

[0009] Performing image enhancement processing on each of the second images using a fast Fourier transform method to obtain a second enhanced image corresponding to each of the second images;

[0010] Reconstructing the first enhanced image and each of the second enhanced images using an inverse discrete wavelet transform reconstruction method to obtain an initial low-light enhanced image;

[0011] The initial low-light enhanced image is enhanced by adopting a structure-weighted guided filtering method to obtain a target low-light enhanced image.

[0012] Optionally, the step of enhancing the first image using an automatic color scale enhancement method to obtain a first enhanced image specifically includes:

[0013] performing a linear transformation on the first image to obtain a third image;

[0014] A linear grayscale stretching process is performed on the third image according to the grayscale parameter of the third image and the predetermined maximum grayscale parameter and the predetermined minimum grayscale parameter of the first enhanced image to obtain the first enhanced image.

[0015] Optionally, performing image enhancement processing on each of the second images using a fast Fourier transform method to obtain a second enhanced image corresponding to each of the second images specifically includes:

[0016] The second image is converted by using a discrete Fourier transform method to obtain a spectrum image and a phase spectrum image;

[0017] performing global contrast enhancement processing on the spectrum image to obtain a spectrum enhanced image;

[0018] Performing edge and texture sharpening and enhancement processing on the phase spectrum image to obtain a phase spectrum enhanced image;

[0019] The second enhanced image is obtained by performing inverse Fourier transform processing based on the spectrum enhanced image and the phase spectrum enhanced image.

[0020] Optionally, the enhancing the initial low-light enhanced image by using a structure-weighted guided filtering method to obtain a target low-light enhanced image specifically includes:

[0021] Decomposing the initial low-light enhanced image to obtain an approximate image and a residual image;

[0022] performing global light adjustment on the approximate image to obtain an enhanced approximate image;

[0023] performing local detail enhancement processing on the residual image to obtain an enhanced residual image;

[0024] A sum operation is performed on the enhanced approximate image and the enhanced residual image to obtain the target low-light enhanced image.

[0025] Optionally, performing global light adjustment on the approximate image to obtain an enhanced approximate image specifically includes:

[0026] performing compression processing on the approximate image to obtain a compressed image;

[0027] Performing calculations based on pixel parameters of the compressed image to obtain a light compression factor;

[0028] Performing calculation based on the target pixel of the compressed image and the light compression factor to obtain a first enhancement factor;

[0029] Global light enhancement processing is performed on the compressed image based on the light compression factor and the first enhancement factor to obtain the enhanced approximate image.

[0030] Optionally, performing calculation processing based on pixel parameters of the compressed image to obtain a light compression factor specifically includes:

[0031] performing mean calculation processing according to pixel parameters of the compressed image to obtain an average pixel value of the compressed image;

[0032] When the average pixel value is greater than or equal to a first preset threshold, performing calculation based on a first difference between a maximum pixel value and a minimum pixel value of the compressed image and the maximum pixel value of the compressed image to obtain the light compression factor;

[0033] When the average pixel value is greater than a second preset threshold and less than the first preset threshold, dividing a second difference between the average pixel value and the minimum pixel value and the first difference to obtain the light compression factor;

[0034] When the average pixel value is less than or equal to the second preset threshold, a division operation is performed based on the average pixel value and the first difference to obtain the light compression factor.

[0035] Optionally, performing local detail enhancement processing on the residual image to obtain an enhanced residual image specifically includes:

[0036] Performing calculation based on the approximate image and the enhanced approximate image to obtain a second enhancement factor;

[0037] Calculation processing is performed based on the second enhancement factor and the residual image to obtain the enhanced residual image.

[0038] To solve the above problems, the present application provides a low-light image adaptive enhancement device, comprising:

[0039] a decomposition module, configured to decompose the to-be-enhanced low-light image using a discrete wavelet transform to obtain sub-band images, wherein the sub-band images include a first image representing low-frequency color features and a plurality of second images representing high-frequency detail features;

[0040] A first enhancement module, configured to enhance the first image using an automatic color scale enhancement method to obtain a first enhanced image;

[0041] a second enhancement module, configured to perform image enhancement processing on each of the second images using a fast Fourier transform method to obtain a second enhanced image corresponding to each of the second images;

[0042] an image reconstruction module, configured to perform image reconstruction based on the first enhanced image and each of the second enhanced images by using an inverse discrete wavelet transform processing method to obtain an initial low-light enhanced image;

[0043] The third enhancement module is configured to perform enhancement processing on the initial low-light enhanced image by adopting a structure-weighted guided filtering method to obtain a target low-light enhanced image.

[0044] To solve the above problem, the present application provides a storage medium, which stores a computer program. When the computer program is executed by a processor, the steps of the above-mentioned low-light image adaptive enhancement method are implemented.

[0045] To solve the above problem, the present application provides an electronic device, which includes at least a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps of the above-mentioned low-light image adaptive enhancement method when executing the computer program in the memory.

[0046] The present application utilizes a discrete wavelet transform (DWT) and a fast Fourier transform (FFT) to decompose the low-light image to be enhanced into different components. The DWT is sensitive to color brightness, while the FFT is sensitive to texture detail. Their combined use effectively enhances the overall color brightness and texture detail of the low-light image, achieving the goal of enhancing the low-light image. The DWT is first applied to the low-light image, decomposing it into different layers of images: a first image representing low-frequency color features and multiple second images representing high-frequency detail features. An automatic color scale enhancement module is then used to enhance the base layer of the first image. A Fourier transform is then used to further enhance the texture detail features of the second image. An adaptive dimming module then fuses the local and global contrast of the low-light image to further improve the brightness and texture detail of the low-light visible image. By combining the DWT and FFT, the present invention enhances low-light images in both the spatial and frequency domains. This not only enhances the overall contrast and weak texture detail of the low-light visible image, but also ensures proper exposure, further improving the quality of the final enhanced low-light image.

[0047] The above description is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are specifically listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Various other advantages and benefits will become apparent to those skilled in the art upon reading the detailed description of the preferred embodiment below. The accompanying drawings are for illustration purposes only and are not to be considered as limiting the present invention. The same reference symbols are used throughout the drawings to represent the same components. In the drawings:

[0049] Figure 1 A schematic diagram of a process for adaptively enhancing a low-light image provided by an embodiment of the present application is shown;

[0050] Figure 2 A schematic diagram of a process for adaptively enhancing a low-light image according to another embodiment of the present application is shown;

[0051] Figure 3 A structural block diagram of a low-light image adaptive enhancement device provided by another embodiment of the present application is shown. DETAILED DESCRIPTION

[0052] Various aspects and features of the present application are described herein with reference to the accompanying drawings.

[0053] It should be understood that various modifications may be made to the embodiments of the present application. Therefore, the above description should not be considered as limiting, but merely as an example of an embodiment. Other modifications within the scope and spirit of the present application will occur to those skilled in the art.

[0054] The accompanying drawings, which are incorporated in and constitute a part of the specification, illustrate embodiments of the present application and, together with the general description of the present application given above and the detailed description of the embodiments given below, serve to explain the principles of the present application.

[0055] These and other characteristics of the present application will become apparent from the following description of a preferred form of embodiment given as a non-limiting example with reference to the accompanying drawings.

[0056] It should also be understood that although the present application has been described with reference to certain specific examples, those skilled in the art will readily be able to implement many other equivalent forms of the present application.

[0057] The above and other aspects, features and advantages of the present application will become more apparent from the following detailed description when taken in conjunction with the accompanying drawings.

[0058] Specific embodiments of the present application will be described hereinafter with reference to the accompanying drawings; however, it should be understood that the embodiments described are merely examples of the present application and may be implemented in a variety of ways. Familiar and / or repetitive functions and structures are not described in detail to avoid obscuring the present application with unnecessary or redundant details. Therefore, the specific structural and functional details described herein are not intended to be limiting, but rather serve merely as a basis and representative basis for the claims to teach those skilled in the art to variously utilize the present application with substantially any suitable detailed structure.

[0059] This specification may use the phrases "in one embodiment," "in another embodiment," "in yet another embodiment," or "in other embodiments," which may all refer to one or more of the same or different embodiments according to the present application.

[0060] The present application embodiment provides a method for adaptively enhancing low-light images. Figure 1 Shown, including:

[0061] Step S101: Decomposing the low-light image to be enhanced by discrete wavelet transform to obtain sub-band images, wherein the sub-band images include a first image representing low-frequency color features and a plurality of second images representing high-frequency detail features;

[0062] In the specific implementation process, Discrete Wavelet Transform (DWT) is used to enhance the low-light image. IDecompose to obtain a sub-band image, wherein the sub-band image includes a first image representing low-frequency color features and multiple second images representing high-frequency detail features; specifically, the first image represents a low-frequency component of a first sub-frequency band that is low-frequency in the horizontal direction and low-frequency in the vertical direction; the second image includes a first high-frequency image representing a second sub-frequency band that is low-frequency in the horizontal direction and high-frequency in the vertical direction , the second high-frequency image of the third sub-band representing high frequency in the horizontal direction and low frequency in the vertical direction and a third high-frequency image of the fourth sub-band representing high frequencies in the horizontal direction and high frequencies in the vertical direction .

[0063] Step S102: performing enhancement processing on the first image using an automatic color scale enhancement method to obtain a first enhanced image;

[0064] In a specific implementation process, a linear transformation process is performed on the first image to obtain a third image; and a linear grayscale stretching process is performed on the third image according to the grayscale parameters of the third image and the predetermined maximum grayscale parameters and predetermined minimum grayscale parameters of the first enhanced image to obtain the first enhanced image.

[0065] Step S103: performing image enhancement processing on each of the second images using a fast Fourier transform method to obtain a second enhanced image corresponding to each of the second images;

[0066] During the specific implementation process, the discrete Fourier transform method (Fast Fourier Transform, abbreviated as FFT) is used to transform the second image to obtain a spectrum image and a phase spectrum image; the spectrum image is subjected to global contrast enhancement processing to obtain a spectrum enhanced image; the phase spectrum image is subjected to edge and texture sharpening enhancement processing to obtain a phase spectrum enhanced image; and the spectrum enhanced image and the phase spectrum enhanced image are subjected to inverse Fourier transform processing to obtain the second enhanced image.

[0067] Step S104: reconstructing the first enhanced image and each of the second enhanced images using an inverse discrete wavelet transform reconstruction method to obtain an initial low-light enhanced image;

[0068] In the specific implementation process, the first enhanced image and each second enhanced image are reconstructed using the inverse discrete wavelet transform (IWT) reconstruction method to obtain an initial low-light enhanced image. .

[0069] Step S105: using a structure-weighted guided filtering method to perform enhancement processing on the initial low-light enhanced image to obtain a target low-light enhanced image.

[0070] During the specific implementation process, the initial low-light enhanced image is decomposed to obtain an approximate image and a residual image; the approximate image is globally adjusted to obtain an enhanced approximate image; the residual image is locally enhanced to obtain an enhanced residual image; and the enhanced approximate image and the enhanced residual image are summed to obtain the target low-light enhanced image.

[0071] This application decomposes the low-light image to be enhanced into different components by employing a discrete wavelet transform and a fast Fourier transform. The discrete wavelet transform is sensitive to color brightness, while the fast Fourier transform is sensitive to texture detail. Their combined use effectively improves the overall color brightness and texture detail of the low-light image, achieving the goal of enhancing the low-light image. The wavelet transform is first introduced to process the low-light image, decomposing it into images at different levels: a first image representing low-frequency color features and multiple second images representing high-frequency detail features. An automatic color scale enhancement module is used to enhance the base layer of the first image, and a Fourier transform is used to further enhance the texture detail features of the second image. An adaptive dimming module then fuses the local and global contrast of the low-light image to further improve the brightness and texture detail of the low-light visible image. By combining wavelet and Fourier transforms, the present invention enhances low-light images in both the spatial and frequency domains. This not only enhances the overall contrast and weak texture detail of the low-light visible image, but also ensures proper exposure of the low-light image, further improving the quality of the final enhanced low-light image.

[0072] Another embodiment of the present application provides another low-light image adaptive enhancement method, such as Figure 2 Shown, including:

[0073] Step S201: Decomposing the low-light image to be enhanced by discrete wavelet transform to obtain sub-band images, wherein the sub-band images include a first image representing low-frequency color features and a plurality of second images representing high-frequency detail features;

[0074] In the specific implementation process of this step, discrete wavelet transform is used to enhance the low-light image. I Decompose to obtain a sub-band image, wherein the sub-band image includes a first image representing low-frequency color features and multiple second images representing high-frequency detail features; specifically, the first image represents a low-frequency component of a first sub-frequency band that is low-frequency in the horizontal direction and low-frequency in the vertical direction; the second image includes a first high-frequency image representing a second sub-frequency band that is low-frequency in the horizontal direction and high-frequency in the vertical direction , the second high-frequency image of the third sub-band representing high frequency in the horizontal direction and low frequency in the vertical direction and a third high-frequency image of the fourth sub-band representing high frequencies in the horizontal direction and high frequencies in the vertical direction Due to its biorthogonality, DWT does not cause image information loss. Given a low-light image I to be enhanced as input, Haar Wavelet Transform is used to decompose the low-light image to be enhanced due to its simplicity and speed. The filters of Haar Wavelet Transform are divided into low-pass filter L and high-pass filter H. The mathematical expression of the low-pass filter can be shown as follows:

[0075]

[0076] The mathematical expression of the high-pass filter can be expressed as follows:

[0077]

[0078] Afterwards, the input low-light image to be enhanced can be transformed into four sub-bands, which can be expressed as follows:

[0079]

[0080] in, I To enhance the low-light image, is the first image that characterizes the low-frequency color features, , , is the second image, which is the first high-frequency image of the second sub-band representing low frequency in the horizontal direction and high frequency in the vertical direction , the second high-frequency image of the third sub-band representing high frequency in the horizontal direction and low frequency in the vertical direction and a third high-frequency image of the fourth sub-band representing high frequencies in the horizontal direction and high frequencies in the vertical direction .

[0081] Step S202: performing linear transformation processing on the first image to obtain a third image;

[0082] In the specific implementation process of this step, the mathematical expression of the linear function of the first image is shown in the following formula (4):

[0083]

[0084] in, Indicates the first image at the two-dimensional coordinate point The pixel value at ; Indicates the third image at the two-dimensional coordinate point The pixel value at ; C and R represent the linear transformation coefficients.

[0085] Step S203: performing linear grayscale stretching on the third image according to the grayscale parameter of the third image and the predetermined maximum grayscale parameter and the predetermined minimum grayscale parameter of the first enhanced image to obtain a first enhanced image. ;

[0086] In the specific implementation of this step, the mathematical expression of linear grayscale stretching can be expressed as the following formula (5):

[0087] (5)

[0088] in, Represents the first enhanced image At two-dimensional coordinate points The pixel value at ; represents the maximum grayscale value of the third image, represents the minimum grayscale value of the third image; Represents the first enhanced image The predetermined maximum grayscale value, Represents the first enhanced image The predetermined minimum grayscale value.

[0089] Step S204: converting the second image using a discrete Fourier transform method to obtain a spectrum image and a phase spectrum image;

[0090] In the specific implementation process of this step, the discrete Fourier transform (FFT) method is used to convert the second image into the Fourier space. The transformation function can be expressed as the following formula (6):

[0091] (6)

[0092] in, Represents the second image in Fourier space The pixel value at ; h, w are the coordinates in two-dimensional space, u, v are the coordinates in Fourier space, j is the imaginary part, Indicates the second image at position coordinates The pixel value at , H represents the image height, and W represents the image width. It can be expressed as the following formula (7):

[0093]

[0094] in, and Fourier space The real and imaginary parts of . Then the spectrum image and phase spectrum image of the second image are obtained; the calculation formula of the spectrum image can be expressed as the following formula (8):

[0095] (8)

[0096] in, Expressed in Fourier space coordinates Spectrum image pixel value at ;

[0097] The calculation formula of the phase spectrum image can be expressed as follows:

[0098]

[0099] in, Expressed in Fourier space coordinates The pixel value of the phase spectrum image at ; The real part of It can also be expressed as the following formula (10):

[0100]

[0101] The imaginary part It can also be expressed as the following formula (11):

[0102]

[0103] Among them, the FFT method is used to , , The feature map is converted into sub-bands, and the first high-frequency image is converted into the first spectrum image using discrete Fourier transform. and the first phase spectrum image , the second high frequency image is converted into the second spectrum image using discrete Fourier transform and the second phase spectrum image , the second high frequency image is converted into the third spectrum image using discrete Fourier transform and the third phase spectrum image .

[0104] Step S205: performing global contrast enhancement processing on the spectrum image to obtain a spectrum enhanced image;

[0105] In the specific implementation of this step, during the spectrum enhancement process, the overall contrast and color performance of the image are enhanced by adjusting the frequency domain energy distribution. The mathematical expression for global contrast enhancement of the spectrum image can be expressed as the following formula (12):

[0106] (12)

[0107] in, is the original spectrum image in Fourier space coordinates The pixel value at Represents the spectrum enhanced image in Fourier space coordinates The pixel value at ; σ is the enhancement strength, which can be greater than or equal to a first threshold and less than or equal to a second threshold; the first threshold can be 1.5, and the second threshold can be 3.0. The enhancement strength can be set according to actual needs. σ is the control enhancement range, which can be 1.8. The control enhancement range can be set according to actual needs.

[0108] Step S206: performing edge and texture sharpening and enhancement processing on the phase spectrum image to obtain a phase spectrum enhanced image;

[0109] In the specific implementation process of this step, during the phase enhancement process, the texture detail feature enhancement is achieved by strengthening the edge and texture information in the phase spectrum. The mathematical expression for edge and texture sharpening enhancement processing of the phase spectrum image can be expressed as the following formula (13):

[0110] (13)

[0111] in, is the Laplace operator, is the sharpening strength, is the original phase spectrum image in Fourier space coordinates The pixel value at Represents the phase spectrum enhanced image in Fourier space coordinates The pixel value at the position; the value of the sharpening intensity is greater than or equal to the third threshold and less than or equal to the fourth threshold; the value of the sharpening intensity can be set according to actual needs.

[0112] Step S207: performing inverse Fourier transform processing based on the spectrum enhanced image and the phase spectrum enhanced image to obtain the second enhanced image;

[0113] In the specific implementation of this step, the mathematical expression of the inverse Fourier transform processing can be expressed as the following formula (14):

[0114] (14)

[0115] in, represents the inverse Fourier transform; e is the exponential operator symbol; j is the imaginary part; It can be expressed as: ;

[0116] Represent the spectrally enhanced image in Fourier space The pixel value at ; Representing the phase spectrum enhanced image in Fourier space The pixel value at ; Represents the second enhanced image obtained by inverse Fourier transform based on the spectrum enhanced image and the phase spectrum enhanced image in Fourier space The pixel value at . The first high frequency image The corresponding second enhanced image is , the second high-frequency image of the third sub-band representing high frequency in the horizontal direction and low frequency in the vertical direction The corresponding second enhanced image is and a third high-frequency image of the fourth sub-band representing high frequencies in the horizontal direction and high frequencies in the vertical direction The corresponding second enhanced image is .

[0117] Step S208: reconstructing the first enhanced image and each of the second enhanced images using an inverse discrete wavelet transform reconstruction method to obtain an initial low-light enhanced image;

[0118] In the specific implementation process of this step, the mathematical expression for reconstructing the first enhanced image and each of the second enhanced images using the discrete wavelet inverse transform reconstruction method can be expressed as the following formula (15):

[0119] (15)

[0120] in, Enhance images for initial low light; is the first enhanced image; The first high-frequency image a corresponding second enhanced image; The second high-frequency image of the third sub-band that represents high frequency in the horizontal direction and low frequency in the vertical direction a corresponding second enhanced image; The third high-frequency image of the fourth sub-band representing high frequency in the horizontal direction and high frequency in the vertical direction The corresponding second enhanced image.

[0121] The inverse discrete wavelet transform (IWT) reconstruction method reconstructs the output of four levels of features to obtain the initial low-light enhanced image: the low-frequency component of the DWT significantly improves color brightness compared to the amplitude component of the FFT, while the phase component of the FFT significantly improves texture detail mapping compared to the high-frequency components of the DWT (LH, HL, and HH). By combining the DWT and FFT, global brightness is enhanced in the low-frequency component, and the phase component of the FFT is used to restore local texture detail and smoothness.

[0122] Step S209: decomposing the initial low-light enhanced image to obtain an approximate image and a residual image;

[0123] In the specific implementation process of this step, the structure-weighted guided filtering (SWGF) is used to decompose the initially enhanced low-light image into an approximate layer and a residual layer for secondary enhancement. The initial low-light enhanced image is decomposed using SWGF to obtain an approximate image and a residual image. The mathematical expression of the approximate image can be expressed as follows:

[0124] (16)

[0125] The mathematical expression of the residual image can be expressed as follows:

[0126]

[0127] in, represents the approximate image output by the approximation layer, Represents the residual image output by the residual layer, set The filter radius is 4, the regularization coefficient is 0.1, Represents the initial low-light enhanced image .

[0128] Step S210: performing global light adjustment on the approximate image to obtain an enhanced approximate image;

[0129] During the specific implementation of this step, the approximate image is compressed to obtain a compressed image. The mathematical formula for the compression process can be expressed as follows:

[0130]

[0131] in, represents a compressed image, represents the natural logarithm operator, A is the approximate image, =1 is a predetermined constant.

[0132] Calculation is performed based on the pixel parameters of the compressed image to obtain a light compression factor; specifically, mean calculation is performed based on the pixel parameters of the compressed image to obtain an average pixel value of the compressed image; when the average pixel value is greater than or equal to a first preset threshold, calculation is performed based on a first difference between a maximum pixel value and a minimum pixel value of the compressed image and the maximum pixel value of the compressed image to obtain the light compression factor; when the average pixel value is greater than a second preset threshold and less than the first preset threshold, division is performed based on a second difference between the average pixel value and the minimum pixel value and the first difference to obtain the light compression factor; when the average pixel value is less than or equal to the second preset threshold, division is performed based on the average pixel value and the first difference to obtain the light compression factor The mathematical expression of the light compression factor can be expressed as follows:

[0133]

[0134] in, , and They are respectively the maximum pixel value, minimum pixel value and average pixel value of the compressed image; the first preset threshold can be 5, the second preset threshold can be 2, and the values ​​of the first preset threshold and the second preset threshold can be set according to actual needs.

[0135] A calculation process is performed based on the target pixel point of the compressed image and the light compression factor to obtain a first enhancement factor; the mathematical expression can be expressed as the following formula (20):

[0136]

[0137] in, is the first enhancement factor, is the light compression factor, and the target pixel is the maximum pixel value of the compressed image .

[0138] The compressed image is subjected to global light enhancement processing based on the light compression factor and the first enhancement factor to obtain the enhanced approximate image. Specifically, the compressed image is subjected to calculation processing based on the light compression factor and the first enhancement factor to obtain an initial enhanced image. The mathematical expression can be expressed as follows:

[0139]

[0140] in, To compress the image, is the initial enhanced approximate image, is the light compression factor, is the first enhancement factor; when When ≤ 1, the contrast of the approximation layer decreases. The initial enhanced approximation image is subjected to overall brightness enhancement processing to obtain the enhanced approximation image. The mathematical expression can be expressed as follows:

[0141]

[0142] in, is the initial enhanced approximate image, To enhance the approximate image, exp represents the natural exponential function operation.

[0143] Step S211: performing local detail enhancement processing on the residual image to obtain an enhanced residual image;

[0144] In the specific implementation process of this step, calculation processing is performed based on the approximate image and the enhanced approximate image to obtain the second enhancement factor; the mathematical expression can be expressed as the following formula (23):

[0145]

[0146] in, represents the second enhancement factor, represents the standard deviation of the image, represents an approximate image, represents the enhanced approximate image after enhancement;

[0147] Based on the second enhancement factor and the residual image Perform calculation processing to obtain the enhanced residual image The mathematical expression can be expressed as follows:

[0148]

[0149] in, represents the residual layer, represents the second enhancement factor, represents the enhanced residual image after enhancement.

[0150] Step S212: performing a sum operation on the enhanced approximate image and the enhanced residual image to obtain the target low-light enhanced image.

[0151] In the specific implementation process of this step, the enhanced approximate image and the enhanced residual image are summed to obtain the target low-light enhanced image. The mathematical expression of the sum of the pixel values ​​at corresponding positions can be expressed as the following formula (25):

[0152]

[0153] in, represents the enhanced approximate image, represents the enhanced residual image, Represents the target low-light enhanced image.

[0154] This application decomposes the low-light image to be enhanced into different components by employing a discrete wavelet transform and a fast Fourier transform. The discrete wavelet transform is sensitive to color brightness, while the fast Fourier transform is sensitive to texture detail. Their combined use effectively improves the overall color brightness and texture detail of the low-light image, achieving the goal of enhancing the low-light image. The wavelet transform is first introduced to process the low-light image, decomposing it into images at different levels: a first image representing low-frequency color features and multiple second images representing high-frequency detail features. An automatic color scale enhancement module is used to enhance the base layer of the first image, and a Fourier transform is used to further enhance the texture detail features of the second image. An adaptive dimming module then fuses the local and global contrast of the low-light image to further improve the brightness and texture detail of the low-light visible image. By combining wavelet and Fourier transforms, the present invention enhances low-light images in both the spatial and frequency domains. This not only enhances the overall contrast and weak texture detail of the low-light visible image, but also ensures proper exposure of the low-light image, further improving the quality of the final enhanced low-light image.

[0155] Another embodiment of the present application provides a low-light image adaptive enhancement device, such as Figure 3 Shown, including:

[0156] Decomposition module 1, configured to decompose the low-light image to be enhanced by using discrete wavelet transform to obtain sub-band images, wherein the sub-band images include a first image representing low-frequency color features and a plurality of second images representing high-frequency detail features;

[0157] A first enhancement module 2, configured to enhance the first image using an automatic color scale enhancement method to obtain a first enhanced image;

[0158] A second enhancement module 3 is configured to perform image enhancement processing on each of the second images using a fast Fourier transform method to obtain a second enhanced image corresponding to each of the second images;

[0159] An image reconstruction module 4 is configured to perform image reconstruction based on the first enhanced image and each of the second enhanced images by using an inverse discrete wavelet transform method to obtain an initial low-light enhanced image;

[0160] The third enhancement module 5 is configured to perform enhancement processing on the initial low-light enhanced image by adopting a structure-weighted guided filtering method to obtain a target low-light enhanced image.

[0161] During the specific implementation process, the first enhancement module 2 is specifically used to: perform linear transformation processing on the first image to obtain a third image; perform linear grayscale stretching processing on the third image according to the grayscale parameters of the third image and the predetermined maximum grayscale parameters and predetermined minimum grayscale parameters of the first enhanced image to obtain the first enhanced image.

[0162] During the specific implementation process, the second enhancement module 3 is specifically used to: use the discrete Fourier transform method to convert the second image to obtain a spectrum image and a phase spectrum image; perform global contrast enhancement processing on the spectrum image to obtain a spectrum enhanced image; perform edge and texture sharpening enhancement processing on the phase spectrum image to obtain a phase spectrum enhanced image; and use inverse Fourier transform processing based on the spectrum enhanced image and the phase spectrum enhanced image to obtain the second enhanced image.

[0163] During the specific implementation process, the third enhancement module 5 is specifically used to: decompose the initial low-light enhanced image to obtain an approximate image and a residual image; perform global light adjustment on the approximate image to obtain an enhanced approximate image; perform local detail enhancement processing on the residual image to obtain an enhanced residual image; and perform a sum operation on the enhanced approximate image and the enhanced residual image to obtain the target low-light enhanced image.

[0164] During the specific implementation process, the third enhancement module 5 is also used to: compress the approximate image to obtain a compressed image; perform calculation processing based on the pixel parameters of the compressed image to obtain a light compression factor; perform calculation processing based on the target pixel points of the compressed image and the light compression factor to obtain a first enhancement factor; perform global light enhancement processing on the compressed image based on the light compression factor and the first enhancement factor to obtain the enhanced approximate image.

[0165] During the specific implementation process, the third enhancement module 5 is also used to: perform mean calculation processing based on the pixel parameters of the compressed image to obtain the average pixel value of the compressed image; when the average pixel value is greater than or equal to a first preset threshold, perform calculation processing based on a first difference between the maximum pixel value and the minimum pixel value of the compressed image and the maximum pixel value of the compressed image to obtain the light compression factor; when the average pixel value is greater than a second preset threshold and less than the first preset threshold, perform division processing based on the second difference between the average pixel value and the minimum pixel value and the first difference to obtain the light compression factor; when the average pixel value is less than or equal to the second preset threshold, perform division processing based on the average pixel value and the first difference to obtain the light compression factor.

[0166] In a specific implementation process, the third enhancement module 5 is further used to: perform calculation processing based on the approximate image and the enhanced approximate image to obtain a second enhancement factor; perform calculation processing based on the second enhancement factor and the residual image to obtain the enhanced residual image.

[0167] This application decomposes the low-light image to be enhanced into different components by employing a discrete wavelet transform and a fast Fourier transform. The discrete wavelet transform is sensitive to color brightness, while the fast Fourier transform is sensitive to texture detail. Their combined use effectively improves the overall color brightness and texture detail of the low-light image, achieving the goal of enhancing the low-light image. The wavelet transform is first introduced to process the low-light image, decomposing it into images at different levels: a first image representing low-frequency color features and multiple second images representing high-frequency detail features. An automatic color scale enhancement module is used to enhance the base layer of the first image, and a Fourier transform is used to further enhance the texture detail features of the second image. An adaptive dimming module then fuses the local and global contrast of the low-light image to further improve the brightness and texture detail of the low-light visible image. By combining wavelet and Fourier transforms, the present invention enhances low-light images in both the spatial and frequency domains. This not only enhances the overall contrast and weak texture detail of the low-light visible image, but also ensures proper exposure of the low-light image, further improving the quality of the final enhanced low-light image.

[0168] Another embodiment of the present application provides a storage medium storing a computer program. When the computer program is executed by a processor, the following method steps are implemented:

[0169] Step 1: Decomposing the low-light image to be enhanced by discrete wavelet transform to obtain sub-band images, wherein the sub-band images include a first image representing low-frequency color features and a plurality of second images representing high-frequency detail features;

[0170] Step 2: using an automatic color scale enhancement method to enhance the first image to obtain a first enhanced image;

[0171] Step 3: Perform image enhancement processing on each of the second images using a fast Fourier transform method to obtain a second enhanced image corresponding to each of the second images;

[0172] Step 4: reconstruct the first enhanced image and each of the second enhanced images using an inverse discrete wavelet transform reconstruction method to obtain an initial low-light enhanced image;

[0173] Step 5: Use a structure-weighted guided filtering method to enhance the initial low-light enhanced image to obtain a target low-light enhanced image.

[0174] Those skilled in the art will understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), Synchronous Link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0175] Those skilled in the art will clearly understand that for the sake of convenience and brevity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0176] The specific implementation process of the above method steps can be found in any of the above embodiments of the low-light image adaptive enhancement method, and will not be repeated in this embodiment.

[0177] This application decomposes the low-light image to be enhanced into different components by employing a discrete wavelet transform and a fast Fourier transform. The discrete wavelet transform is sensitive to color brightness, while the fast Fourier transform is sensitive to texture detail. Their combined use effectively improves the overall color brightness and texture detail of the low-light image, achieving the goal of enhancing the low-light image. The wavelet transform is first introduced to process the low-light image, decomposing it into images at different levels: a first image representing low-frequency color features and multiple second images representing high-frequency detail features. An automatic color scale enhancement module is used to enhance the base layer of the first image, and a Fourier transform is used to further enhance the texture detail features of the second image. An adaptive dimming module then fuses the local and global contrast of the low-light image to further improve the brightness and texture detail of the low-light visible image. By combining wavelet and Fourier transforms, the present invention enhances low-light images in both the spatial and frequency domains. This not only enhances the overall contrast and weak texture detail of the low-light visible image, but also ensures proper exposure of the low-light image, further improving the quality of the final enhanced low-light image.

[0178] Another embodiment of the present application provides an electronic device, which may be a server, and the electronic device includes a processor, a memory, a network interface, and a database connected via a system bus. The processor of the electronic device is used to provide computing and control capabilities. The memory of the electronic device includes a non-volatile and / or volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the electronic device is used to communicate with an external client via a network connection. When the electronic device program is executed by the processor, it implements the functions or steps of a server-side method for adaptively enhancing low-light images.

[0179] In one embodiment, an electronic device is provided, which may be a client. The electronic device includes a processor, a memory, a network interface, a display screen, and an input device connected via a system bus. The processor of the electronic device is used to provide computing and control capabilities. The memory of the electronic device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The network interface of the electronic device is used to communicate with an external server via a network connection. When the electronic device program is executed by the processor, it implements the functions or steps on the client side of a method for adaptively enhancing low-light images.

[0180] Another embodiment of the present application provides an electronic device, comprising at least a memory and a processor, wherein the memory stores a computer program, and the processor implements the following method steps when executing the computer program in the memory:

[0181] Step 1: Decomposing the low-light image to be enhanced by discrete wavelet transform to obtain sub-band images, wherein the sub-band images include a first image representing low-frequency color features and a plurality of second images representing high-frequency detail features;

[0182] Step 2: using an automatic color scale enhancement method to enhance the first image to obtain a first enhanced image;

[0183] Step 3: Perform image enhancement processing on each of the second images using a fast Fourier transform method to obtain a second enhanced image corresponding to each of the second images;

[0184] Step 4: reconstruct the first enhanced image and each of the second enhanced images using an inverse discrete wavelet transform reconstruction method to obtain an initial low-light enhanced image;

[0185] Step 5: Use a structure-weighted guided filtering method to enhance the initial low-light enhanced image to obtain a target low-light enhanced image.

[0186] The specific implementation process of the above method steps can be found in any of the above embodiments of the low-light image adaptive enhancement method, and will not be repeated in this embodiment.

[0187] This application decomposes the low-light image to be enhanced into different components by employing a discrete wavelet transform and a fast Fourier transform. The discrete wavelet transform is sensitive to color brightness, while the fast Fourier transform is sensitive to texture detail. Their combined use effectively improves the overall color brightness and texture detail of the low-light image, achieving the goal of enhancing the low-light image. The wavelet transform is first introduced to process the low-light image, decomposing it into images at different levels: a first image representing low-frequency color features and multiple second images representing high-frequency detail features. An automatic color scale enhancement module is used to enhance the base layer of the first image, and a Fourier transform is used to further enhance the texture detail features of the second image. An adaptive dimming module then fuses the local and global contrast of the low-light image to further improve the brightness and texture detail of the low-light visible image. By combining wavelet and Fourier transforms, the present invention enhances low-light images in both the spatial and frequency domains. This not only enhances the overall contrast and weak texture detail of the low-light visible image, but also ensures proper exposure of the low-light image, further improving the quality of the final enhanced low-light image.

[0188] The above embodiments are merely exemplary embodiments of the present application and are not intended to limit the scope of the present application. The scope of protection of the present application is defined by the claims. Those skilled in the art may make various modifications or equivalent substitutions to the present application within the essence and scope of protection of the present application, and such modifications or equivalent substitutions shall also be deemed to fall within the scope of protection of the present application.

Claims

1. A method for adaptively enhancing low-light images, characterized in that: include: Decomposing the low-light image to be enhanced by discrete wavelet transform to obtain sub-band images, wherein the sub-band images include a first image representing low-frequency color features and a plurality of second images representing high-frequency detail features; Performing enhancement processing on the first image using an automatic color scale enhancement method to obtain a first enhanced image; Performing image enhancement processing on each of the second images using a fast Fourier transform method to obtain a second enhanced image corresponding to each of the second images; Reconstructing the first enhanced image and each of the second enhanced images using an inverse discrete wavelet transform reconstruction method to obtain an initial low-light enhanced image; Using a structure-weighted guided filtering method to enhance the initial low-light enhanced image to obtain a target low-light enhanced image; The step of enhancing the first image by using the automatic color scale enhancement method to obtain a first enhanced image specifically includes: performing a linear transformation on the first image to obtain a third image; performing linear grayscale stretching processing on the third image according to the grayscale parameter of the third image and the predetermined maximum grayscale parameter and the predetermined minimum grayscale parameter of the first enhanced image to obtain the first enhanced image; The method of performing image enhancement processing on each of the second images using a fast Fourier transform method to obtain a second enhanced image corresponding to each of the second images specifically includes: Converting the second image using a discrete Fourier transform method to obtain a spectrum image and a phase spectrum image; performing global contrast enhancement processing on the spectrum image to obtain a spectrum enhanced image; Performing edge and texture sharpening and enhancement processing on the phase spectrum image to obtain a phase spectrum enhanced image; The second enhanced image is obtained by performing inverse Fourier transform processing based on the spectrum enhanced image and the phase spectrum enhanced image.

2. The method according to claim 1, wherein The method of enhancing the initial low-light enhanced image by using a structure-weighted guided filtering method to obtain a target low-light enhanced image specifically includes: Decomposing the initial low-light enhanced image to obtain an approximate image and a residual image; performing global light adjustment on the approximate image to obtain an enhanced approximate image; performing local detail enhancement processing on the residual image to obtain an enhanced residual image; A sum operation is performed on the enhanced approximate image and the enhanced residual image to obtain the target low-light enhanced image.

3. The method according to claim 2, wherein The performing global light adjustment on the approximate image to obtain an enhanced approximate image specifically includes: performing compression processing on the approximate image to obtain a compressed image; Performing calculations based on pixel parameters of the compressed image to obtain a light compression factor; Performing calculation based on the target pixel of the compressed image and the light compression factor to obtain a first enhancement factor; Global light enhancement processing is performed on the compressed image based on the light compression factor and the first enhancement factor to obtain the enhanced approximate image.

4. The method according to claim 3, wherein The calculating and processing according to the pixel parameters of the compressed image to obtain the light compression factor specifically includes: performing mean calculation processing according to pixel parameters of the compressed image to obtain an average pixel value of the compressed image; When the average pixel value is greater than or equal to a first preset threshold, performing calculation based on a first difference between a maximum pixel value and a minimum pixel value of the compressed image and the maximum pixel value of the compressed image to obtain the light compression factor; When the average pixel value is greater than a second preset threshold and less than the first preset threshold, dividing a second difference between the average pixel value and the minimum pixel value and the first difference to obtain the light compression factor; When the average pixel value is less than or equal to the second preset threshold, a division operation is performed based on the average pixel value and the first difference to obtain the light compression factor.

5. The method according to claim 2, wherein The performing local detail enhancement processing on the residual image to obtain an enhanced residual image specifically includes: Performing calculation based on the approximate image and the enhanced approximate image to obtain a second enhancement factor; Calculation processing is performed based on the second enhancement factor and the residual image to obtain the enhanced residual image.

6. A low-light image adaptive enhancement device, characterized in that: include: a decomposition module, configured to decompose the to-be-enhanced low-light image using a discrete wavelet transform to obtain sub-band images, wherein the sub-band images include a first image representing low-frequency color features and a plurality of second images representing high-frequency detail features; a first enhancement module, configured to enhance the first image using an automatic color scale enhancement method to obtain a first enhanced image; specifically, to perform a linear transformation on the first image to obtain a third image; and to perform a linear grayscale stretching on the third image based on a grayscale parameter of the third image and a predetermined maximum grayscale parameter and a predetermined minimum grayscale parameter of the first enhanced image to obtain the first enhanced image; a second enhancement module, configured to perform image enhancement processing on each of the second images using a fast Fourier transform method to obtain a second enhanced image corresponding to each of the second images; and specifically configured to convert the second images using a discrete Fourier transform method to obtain a spectrum image and a phase spectrum image; performing global contrast enhancement processing on the spectrum image to obtain a spectrum enhanced image; Performing edge and texture sharpening and enhancement processing on the phase spectrum image to obtain a phase spectrum enhanced image; Obtaining the second enhanced image by performing inverse Fourier transform processing on the spectrum enhanced image and the phase spectrum enhanced image; an image reconstruction module, configured to perform image reconstruction based on the first enhanced image and each of the second enhanced images by using an inverse discrete wavelet transform method to obtain an initial low-light enhanced image; The third enhancement module is configured to perform enhancement processing on the initial low-light enhanced image by adopting a structure-weighted guided filtering method to obtain a target low-light enhanced image.

7. A storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by the processor, the steps of the low-light image adaptive enhancement method according to any one of claims 1 to 5 are implemented.

8. An electronic device, characterized in that: The method comprises at least a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program in the memory, the method implements the steps of the method for adaptively enhancing low-light images according to any one of claims 1 to 5.

Images