A contrast enhancement method of remote sensing image

Abstract

The application discloses a remote sensing image contrast enhancement method, and belongs to the technical field of remote sensing image processing. The method comprises the following steps: obtaining a first gray threshold of a remote sensing image, and dividing the remote sensing image into a saturation region, a foreground region and a background region; constructing a first optimization model, and configuring a first transmittance map of the foreground region by using the first optimization model; configuring a second transmittance map of the background region, and reducing pixel gray values of the second transmittance map in a second gray interval to obtain a third transmittance map of the saturation region; and merging the third transmittance map with the adjusted first transmittance map and the second transmittance map to obtain a contrast enhancement image. The first optimization model is constructed to solve the foreground region, so that the original gray level relationship is maintained while the contrast is improved. The second transmittance map of the background region is adjusted in gray, so that the background details are enhanced and the interference on the foreground region is avoided.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of remote sensing image processing, and particularly relates to a contrast enhancement method of remote sensing image. BACKGROUND

[0002] At present, in the field of remote sensing image enhancement, methods based on atmospheric scattering model are widely used. Such methods quantize the image degradation process as two key parameters of atmospheric light and transmittance by constructing a physical model of light transmission in atmospheric medium, and realize accurate recovery of scene information. However, when applied to single-channel remote sensing images, the existing atmospheric scattering model method still has obvious limitations. Since the single-channel image itself lacks color information, the traditional transmittance estimation mechanism is prone to estimation deviation or even failure in high-light areas, and it is difficult to adapt to the complex and variable dynamic range distribution in real remote sensing scenes. This deficiency leads to a dilemma in enhancement processing: if the high-light area details are prioritized, the contrast of the shadow area is easily suppressed; if the overall visual performance is improved, the original gray level and structure information of the image may be damaged, affecting the accuracy of subsequent analysis and application of remote sensing images. SUMMARY

[0003] Therefore, the present application aims to provide a contrast enhancement method of remote sensing image, which automatically identifies the remote sensing image into saturated area, foreground area and background area through intelligent region division technology, solves the foreground area by constructing a first optimization model to ensure that the original gray level relationship is maintained while the contrast is improved, and adjusts the second transmittance map of the background area through empirical parameters, which not only enhances the background details, but also avoids interference with the foreground area, and ensures the original gray level and structure information of the remote sensing image.

[0004] To achieve the above purpose, the technical scheme of the present application is as follows:

[0005] The present application provides a contrast enhancement method of remote sensing image, comprising:

[0006] S1: obtaining a first gray threshold of the remote sensing image, and dividing the remote sensing image into a saturated area and an under-enhanced area according to the first gray threshold;

[0007] S2: obtaining a pixel gray value of the remote sensing image, and dividing the under-enhanced area into a foreground area and a background area according to the pixel gray value, wherein the foreground area is provided with a first gray interval, the background area is provided with a second gray interval, and the lower limit of the first gray interval is higher than the upper limit of the second gray interval;

[0008] S3: a first optimization model is constructed, and the first optimization model is used to configure a first transmittance map of the foreground region, the first optimization model is used to enhance pixel gray values of the first transmittance map in a first gray interval; a second gray threshold of the background region and a pixel gray mean value of the background region are obtained, a second transmittance map of the background region is configured according to the second gray threshold and the pixel gray mean value, and pixel gray values of the second transmittance map are reduced in a second gray interval; a transmittance of the saturated region is set to 1 to obtain a third transmittance map of the saturated region;

[0009] S4: the third transmittance map is merged with the adjusted first transmittance map and the second transmittance map to obtain a contrast-enhanced image of the remote sensing image.

[0010] Preferably, the first optimization model is constructed, and the first optimization model is used to configure the first transmittance map of the foreground region, comprising: S31: a degraded image of the remote sensing image is obtained, and a bright channel value, a dark channel value and an auxiliary parameter of the degraded image are determined;

[0011] S32: a transmittance lower limit value of the foreground region is calculated according to the bright channel value, the dark channel value and the auxiliary parameter;

[0012] S33: a mask matrix is constructed according to a preset Gaussian kernel and the degraded image;

[0013] S34: a gradient of the first transmittance map is obtained, the first optimization model is constructed according to the gradient of the first transmittance map, the mask matrix and the transmittance lower limit value, and a preset first regularization parameter, a second regularization parameter and a transmittance upper limit value, the foreground region is solved by using the first optimization model to obtain an optimized transmittance vector of the foreground region, and the first transmittance map is obtained according to the optimized transmittance vector.

[0014] Preferably, the transmittance lower limit value of the foreground region is calculated according to the bright channel value, the dark channel value and the auxiliary parameter, and is represented by the following formula:

[0015] ;

[0016] Wherein, is the transmittance lower limit value, is a neighborhood pixel set, is a pixel point in the neighborhood pixel set, is the bright channel value, is the dark channel value, is the auxiliary parameter.

[0017] Preferably, the first optimization model is constructed, and is represented by the following formula:

[0018] ;

[0019] wherein, is a first regularization parameter, is a second regularization parameter, is a first transmittance map, is a transmittance lower limit value, is a preset transmittance upper limit value, is a mask matrix, is a gradient of the first transmittance map, is a smoothing constraint term, is a norm regularization parameter.

[0020] Preferably, the first optimization model is used to solve the foreground region to obtain an optimized transmittance vector of the foreground region, including:

[0021] S341: converting the first optimization model into a second optimization model, and solving the foreground region according to the second optimization model to obtain a gradient transmittance map of the foreground region;

[0022] S342: converting the mask matrix into a sparse matrix, converting the transmittance upper limit value into a transmittance upper limit vector, converting the transmittance lower limit value into a transmittance lower limit vector, and converting the gradient transmittance map into a first gradient transmittance vector and a second gradient transmittance vector;

[0023] S343: obtaining preset first gradient sparse operators and second gradient sparse operators, and obtaining an optimized transmittance vector according to the sparse matrix, the transmittance upper limit vector, the transmittance lower limit vector, the first gradient transmittance vector, the second gradient transmittance vector, the first gradient sparse operators and the second gradient sparse operators.

[0024] Preferably, the second transmittance map of the background region is configured according to the second gray threshold and the pixel gray mean value, including:

[0025] S35: obtaining all pixel gray values in the background region, and obtaining a pixel gray mean value of the image block according to the pixel gray values;

[0026] S36: correcting a difference between the gray median and the pixel gray mean value according to a preset empirical parameter, adjusting the pixel gray values of the image block of the background region, and obtaining the second transmittance map of the background region.

[0027] Preferably, the difference between the gray median and the pixel gray mean value is corrected according to a preset empirical parameter, and is represented by the following formula:

[0028] ;

[0029] wherein, is a second transmittance map, is an empirical parameter, is a gray median, is the average gray value of the pixel.

[0030] Preferably, the third transmittance map is combined with the adjusted first transmittance map and the second transmittance map to obtain a contrast-enhanced image of the remote sensing image, comprising:

[0031] S41: obtaining a degraded image of the remote sensing image, and obtaining an atmospheric light component of the degraded image;

[0032] S42: obtaining a contrast-enhanced image according to the degraded image, the atmospheric light component, the first transmittance map, the second transmittance map and the third transmittance map.

[0033] The remote sensing image contrast enhancement method according to the embodiment of the present application has at least the following beneficial effects:

[0034] By automatically identifying the remote sensing image into a saturated region, a foreground region and a background region, for the foreground region, a first optimization model is constructed to solve the foreground region under the premise of strictly following the ASM physical constraint, to obtain a first transmittance map, to ensure that the original gray level relationship of the remote sensing image is maintained while the contrast of the remote sensing image is improved. The second transmittance map of the background region is adjusted by an empirical parameter, which not only enhances the details of the background region, but also avoids interference with the foreground region, and ensures the original gray level and structure information of the remote sensing image. In addition, the remote sensing image contrast enhancement process adopts an efficient numerical optimization method, which not only ensures the processing effect of the remote sensing image, but also meets the real-time requirement in actual application process.

[0035] Further, the embodiment effectively solves the technical problem of uneven contrast distribution in single-channel remote sensing images by organically combining the non-uniformity enhancement of the remote sensing image with the atmospheric scattering model, through the synergistic effect of physical constraint and intelligent optimization. Compared with the prior art, the embodiment can significantly improve the detail performance of the remote sensing image after contrast enhancement under the premise of maintaining the naturalness of the image, especially in application scenarios such as target detection and ground feature classification, which have high requirements for image quality, and can improve the data basis of higher quality for subsequent analysis and application of remote sensing images. BRIEF DESCRIPTION OF DRAWINGS

[0036] The accompanying drawings, which form a part of the present application, are included to provide a further understanding of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the present application and together with the description serve to explain the present application. In the drawings:

[0037] Figure 1 is a flowchart of a remote sensing image contrast enhancement method;

[0038] Figure 2This is a flowchart of constructing the first optimization model and using the first optimization model to calculate the first transmittance map of the foreground region;

[0039] Figure 3 The flowchart shows how the first optimization model solves for the background region to obtain the optimized transmittance vector of the background region.

[0040] Figure 4 This is a flowchart for calculating the second transmittance map of the background area based on the grayscale threshold and the average grayscale value of the pixels;

[0041] Figure 5 This is a flowchart for obtaining a contrast-enhanced image of a remotely sensed image;

[0042] Figure 6 These are remote sensing images provided in embodiments of the present invention;

[0043] Figure 7 This is a transmittance map of a remote sensing image provided in an embodiment of the present invention;

[0044] Figure 8 This is an image of the foreground region provided in an embodiment of the present invention;

[0045] Figure 9 This is a first transmittance map of the foreground region provided in an embodiment of the present invention;

[0046] Figure 10 This is an image of the background area provided in the embodiments of the present invention;

[0047] Figure 11 This is a second transmittance map of the background area provided in an embodiment of the present invention;

[0048] Figure 12 This is an image of the saturated region provided in the embodiments of the present invention;

[0049] Figure 13 This is the third transmittance map of the saturation region provided in the embodiments of the present invention;

[0050] Figure 14 It is a contrast-enhanced image of a remote sensing image provided in the embodiments of the present invention;

[0051] Figure 15 This is a transmittance map of a contrast-enhanced image provided in an embodiment of the present invention. Detailed Implementation

[0052] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only for explaining the invention and do not constitute a limitation thereof. Similar elements in different embodiments are referred to by associated similar element reference numerals. In the following embodiments, many details are described to facilitate a better understanding of the invention. However, those skilled in the art will readily recognize that some features may be omitted in different situations, or may be replaced by other elements, materials, or methods. In some cases, some operations related to the invention are not shown or described in the specification. This is to avoid obscuring the core parts of the invention with excessive description. For those skilled in the art, detailed description of these related operations is not necessary; they can fully understand the related operations based on the description in the specification and general technical knowledge in the art.

[0053] It should be noted that, unless otherwise specified, the embodiments and features described in this invention can be combined to form various implementations. Furthermore, the order of the steps or actions in the method description can be changed or adjusted in a manner readily apparent to those skilled in the art. Therefore, the various orders in the specification and drawings are merely for the clear description of a particular embodiment and do not imply a mandatory order, unless otherwise stated that a particular order must be followed.

[0054] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicating orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation on this invention. Furthermore, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, features defined with "first," "second," etc., may explicitly or implicitly include one or more of that feature. In the description of this invention, unless otherwise stated, "a plurality of" means two or more.

[0055] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art will understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0056] The invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0057] Reference Figure 1 , Figure 1 This is an overall flowchart of the remote sensing image contrast enhancement method of the present invention, which includes the following steps:

[0058] S1: Obtain the first grayscale threshold of the remote sensing image, and divide the remote sensing image into saturated regions and under-enhanced regions according to the first grayscale threshold;

[0059] S2: Obtain the pixel grayscale values ​​of the remote sensing image, and divide the under-enhanced area into a foreground area and a background area according to the pixel grayscale values. The foreground area is set with a first grayscale range, and the background area is set with a second grayscale range. The lower limit of the first grayscale range is higher than the upper limit of the second grayscale range.

[0060] S3: Construct a first optimization model and use it to configure a first transmittance map of the foreground region. The first optimization model is used to enhance the pixel grayscale values ​​of the first transmittance map within a first grayscale range. Obtain a second grayscale threshold and the average pixel grayscale value of the background region. Configure a second transmittance map of the background region based on the second grayscale threshold and the average pixel grayscale value, and reduce the pixel grayscale values ​​of the second transmittance map within a second grayscale range. Set the transmittance of the saturated region to 1 to obtain a third transmittance map of the saturated region.

[0061] S4: Merge the third transmittance map with the adjusted first and second transmittance maps to obtain a contrast-enhanced image of the remote sensing image.

[0062] It should be noted that, based on the pixel grayscale value and the set of neighboring pixels, the remote sensing image is divided into saturated regions and under-enhanced regions. Then, by using the median grayscale value of the image patch in the acquired remote sensing image, the under-enhanced region is further divided into foreground and background regions.

[0063] Furthermore, the foreground region is solved using the first optimization model to obtain a first transmittance map of the foreground region. The pixel grayscale values ​​in the first transmittance map are then enhanced to varying degrees. A second transmittance map of the background region is obtained using the median and mean grayscale values. Within the range of pixel grayscale values ​​in the background region, the grayscale characteristics of each image block in the background region are optimized to ensure that the first transmittance map of the foreground region has sufficient grayscale representation space. In this embodiment, since the pixel grayscale values ​​in the saturated region are already within the maximum dynamic range, the saturated region of the remote sensing image is not enhanced to avoid over-enhancement leading to loss of detail in the remote sensing image. Therefore, a third transmittance map of the saturated region is directly obtained and assigned a value of 1, i.e., ,in, The third transmittance map is used as the adjustment standard to enhance the pixel grayscale values ​​of the first transmittance map so that the pixel grayscale values ​​of the first transmittance map are close to those of the third transmittance map in the saturation region. Then, the pixel grayscale values ​​of the second transmittance map are adaptively adjusted to enhance the contrast difference between the foreground and background regions.

[0064] Furthermore, this embodiment sets the pixel grayscale values ​​of the background and foreground regions non-linearly, so that the overall grayscale relationship between the saturated region, the foreground region, and the background region remains unchanged, and ensures that the dynamic range distribution of the image after contrast enhancement is more reasonable.

[0065] The pixel grayscale values ​​of the background and foreground regions are set non-linearly. The specific implementation process is as follows:

[0066] For the foreground region, the first optimization model is used to solve the foreground region to ensure that in the first transmittance map of the foreground region, the increase in pixel gray value is relatively small for image blocks with large pixel gray value, and the increase in pixel gray value is relatively large for image blocks with small pixel gray value. Through the above mechanism, the non-linear setting of the foreground region is realized, thereby significantly enhancing the contrast in the foreground region, while avoiding the chaos of gray value relationship of image blocks.

[0067] For the background region, the pixel grayscale value of each image block in the second transmittance map is optimized and adjusted according to the pixel grayscale range. That is, the pixel grayscale value of different image blocks in the background region is reduced, which not only enhances the contrast of the background region, but also maintains the distinction and grayscale range between the background region and the foreground region. This significantly improves the contrast difference between the foreground region and the background region, and further enhances the adaptability and non-uniformity of the remote sensing image when performing contrast enhancement.

[0068] It should be noted that the remote sensing images in this embodiment are all single-channel remote sensing images. Each pixel in a single-channel remote sensing image contains only one value, which in this embodiment is the pixel grayscale value.

[0069] It should be noted that, referring to Figure 6 and Figure 7 As can be seen, the initial remote sensing image is rather blurry, and the contrast difference is not clear. Therefore, in this embodiment, the remote sensing image is first divided into saturation regions (such as pixel grayscale values, neighboring pixel sets, and median grayscale values). Figure 8 As shown), foreground area (such as) Figure 10 (as shown) and background area (such as) Figure 12 As shown), therefore, by constructing the first optimization model, the foreground region is solved to obtain the first transmittance map of the foreground region (as shown). Figure 9 As shown), obtain the gray-level mean of the image patch, and obtain the second transmittance map of the background region based on the gray-level median and gray-level mean (e.g., Figure 11 As shown), obtain the third transmittance map of the saturated region (e.g. Figure 13 As shown), a value assignment operation is performed on the third transmittance map. Based on the third transmittance map after the assignment operation, the pixel gray values ​​of the foreground region and the background region of the first and second transmittance maps are adjusted. The first, second, and third transmittance maps are then merged to obtain a contrast-enhanced image of the remote sensing image (e.g., Figure 14 (as shown), refer to Figure 14 and Figure 15 As can be seen, the contrast-enhanced image in this embodiment has a high contrast, which can effectively preserve the original grayscale levels and structural information of the remote sensing image, and significantly improve the detail performance of the contrast-enhanced remote sensing image while ensuring the naturalness of the remote sensing image.

[0070] It should be noted that, Figure 7 , Figure 9 , Figure 11 , Figure 13 and Figure 15 The X and Y axes represent the pixel coordinates of the remote sensing image, and the Z axis represents the transmittance value of the remote sensing image.

[0071] It should be noted that regions where the set of neighboring pixels is greater than or equal to the pixel threshold are considered saturated regions, while regions where the set of neighboring pixels is less than the pixel threshold are considered under-enhanced regions, as expressed by the following first formula:

[0072] ;

[0073] in, For the set of neighboring pixels, For pixels that are within the neighboring pixel set, For pixel threshold, For image blocks, For saturated regions, Foreground area, For background area, Enhancement algorithms for solving foreground regions Enhancement algorithms for solving background regions This is the pixel threshold.

[0074] Additionally, in one embodiment, reference is made to Figure 2 ,exist Figure 1 In step S3 of the illustrated embodiment, constructing a first optimization model to solve for the foreground region also includes the following steps:

[0075] S31: Acquire the degraded image of the remote sensing image and determine the bright channel value, dark channel value and auxiliary parameters of the degraded image;

[0076] S32: Calculate the lower limit of transmittance of the foreground region based on the bright channel value, dark channel value, and auxiliary parameters;

[0077] S33: Construct a mask matrix based on the preset Gaussian kernel and the degraded image;

[0078] S34: Obtain the gradient of the first transmittance map, construct a first optimization model based on the mask matrix, the lower limit of transmittance and the gradient of the first transmittance map, as well as the preset regularization parameters and the upper limit of transmittance, solve the foreground region using the first optimization model to obtain the optimized transmittance vector of the foreground region, and obtain the first transmittance map based on the optimized transmittance vector.

[0079] It should be noted that the lower limit of transmittance of the foreground region is calculated based on the bright channel value, dark channel value, and auxiliary parameters, and is expressed by the following second formula:

[0080] ;

[0081] in, Lower limit of transmittance For the set of neighboring pixels, For pixels that are within the neighboring pixel set, To highlight channel values, For dark channel value, These are auxiliary parameters.

[0082] It should be noted that auxiliary parameters This is a relatively small value used to increase the overall value of the lower limit of transmittance, thereby improving the robustness of the first optimization model. In this embodiment, the auxiliary parameter... Set it to 0.1.

[0083] The foreground region is solved using the first optimization model to obtain the first transmittance map, and the first transmittance map is positioned... Within the range, This is the lower limit of transmittance. This is the first transmittance map. To ensure contrast enhancement of the remote sensing image, the increase in pixel grayscale value is smaller for image blocks with larger grayscale values ​​in the first transmittance map, and larger for image blocks with smaller grayscale values.

[0084] Furthermore, the mask matrix constructed based on the preset Gaussian kernel and the degraded image is expressed by the following third formula:

[0085] ;

[0086] in, For mask matrix, For Gaussian kernel, This is a degraded image.

[0087] It should be noted that the mask matrix is ​​used to determine how close the pixel grayscale values ​​of image patches in different regions of the first transmittance map are to the saturated region. By defining the mask matrix as the convolution of the degraded image with a Gaussian kernel, the distinguishability of image patches in the first transmittance map is enhanced.

[0088] It should be noted that the first optimization model is constructed based on the mask matrix, the lower limit of transmittance, and the preset regularization parameters and upper limit of transmittance, and is expressed by the following fourth formula:

[0089] ;

[0090] in, For the first regularization parameter, For the second regularization parameter, For the first transmittance diagram, Lower limit of transmittance The preset upper limit of transmittance, For mask matrix, The gradient of the first transmittance map, To smooth out constraint terms, The norm regularization parameter.

[0091] It should be noted that the smoothing constraint term This is used to ensure local similarity between different pixel grayscale values ​​in the first transmittance map, thereby avoiding artifacts during the foreground region contrast enhancement process.

[0092] Those skilled in the art will understand that, since the first optimization model optimizes the image of the foreground region, it is not necessary to introduce the coordinates of the first transmittance map. Therefore, in the fourth formula, the coordinates of the second formula mentioned above are not required. of and of The formulas are omitted, and the same applies to subsequent formulas.

[0093] Additionally, in one embodiment, reference is made to Figure 3 ,exist Figure 2 In step S34 of the illustrated embodiment, the first optimization model solves for the foreground region to obtain the optimized transmittance vector of the foreground region, and further includes the following steps:

[0094] S341: Convert the first optimization model into the second optimization model, solve the foreground region according to the second optimization model, and obtain the gradient transmittance map of the foreground region;

[0095] S342: Convert the mask matrix into a sparse matrix; convert the upper limit of transmittance into an upper limit vector of transmittance; convert the lower limit of transmittance into a lower limit vector of transmittance; convert the gradient transmittance map into a first gradient transmittance vector and a second gradient transmittance vector.

[0096] S343: Obtain the preset first gradient sparse operator and second gradient sparse operator, and obtain the optimized transmittance vector based on the sparse matrix, transmittance upper limit vector, transmittance lower limit vector, first gradient transmittance vector, second gradient transmittance vector, first gradient sparse operator and second gradient sparse operator.

[0097] It should be noted that auxiliary variables are introduced when solving the first optimization model described above. and order The first optimization model is transformed into the second optimization model, as expressed by the following fifth formula:

[0098] ;

[0099] in, For the first regularization parameter, For the second regularization parameter, For the third regularization parameter, For the first transmittance diagram, Lower limit of transmittance The preset upper limit of transmittance, For mask matrix, The gradient of the first transmittance map, To smooth out constraint terms, For auxiliary variables, The norm regularization parameter.

[0100] It should be noted that, in order to efficiently solve the first transmittance map, this embodiment introduces an auxiliary variable. The gradient of the approximate first transmittance map This decomposes the original problem into two subproblems that can be solved iteratively and alternately: an auxiliary variable The solution involves one aspect: the first transmittance map. The above method significantly reduces the computational complexity of solving the first transmittance map while ensuring that the first transmittance map has good edge preservation and smoothness properties.

[0101] Furthermore, auxiliary variables The solution is expressed by the following sixth formula:

[0102] ;

[0103] in, For the auxiliary variables after iteration, The gradient of the first transmittance map after iteration, To smooth out constraint terms, For the third regularization parameter, For the number of iterations, Otherwise, The norm regularization parameter.

[0104] It should be noted that, firstly, the auxiliary variables are updated based on the gradient of the first transmittance map after iteration. Only gradients greater than the ratio of the smoothing constraint line to the third regularization parameter are retained, while smaller gradients are set to zero, thereby strengthening the constraint on the gradient sparsity of the first transmittance map.

[0105] By introducing auxiliary variables to alternately optimize the pixel grayscale values ​​in the first transmittance map, the efficiency and robustness of the algorithm are ensured, so that the final first transmittance map can maintain sharp edges and avoid artifacts in flat areas of the remote sensing image.

[0106] Furthermore, the first transmittance map is solved to transform it into an optimized transmittance vector. Based on the sparse matrix, the upper transmittance bound vector, the lower transmittance bound vector, the first gradient transmittance vector, the second gradient transmittance vector, the first gradient sparse operator, and the second gradient sparse operator, the optimized transmittance vector is obtained, expressed by the following seventh formula:

[0107] ;

[0108] in, To optimize the transmittance vector, For the first regularization parameter, For the second regularization parameter, For the third regularization parameter, For transpose operation, For mask matrix, For sparse matrices, For the upper limit vector of transmittance, For the lower limit vector of transmittance, The first gradient transmittance vector, Let P be the second gradient transmittance vector and P be the normalization term.

[0109] Additionally, in one embodiment, reference is made to Figure 4 ,exist Figure 1 In step S3 of the illustrated embodiment, each image block is optimized to obtain a second transmittance map of the background region, and the following steps are also included:

[0110] S35: Obtain the pixel grayscale values ​​of all image blocks within the background area, and obtain the average pixel grayscale value of the image blocks based on the pixel grayscale values;

[0111] S36: Correct the difference between the median gray level and the mean gray level of the pixels according to the preset empirical parameters, so as to adjust the pixel gray level of the image block in the background area and obtain the second transmittance map of the background area.

[0112] It should be noted that the difference between the median grayscale value and the mean grayscale value is corrected based on preset empirical parameters, as expressed by the following eighth formula:

[0113] ;

[0114] in, For the second transmittance map, For empirical parameters, The median of grayscale This represents the average grayscale value of the pixels.

[0115] It should be noted that, in this embodiment, by adjusting the grayscale value of the second transmittance map to different degrees according to empirical parameters in different image blocks, the contrast of the background area is enhanced while maintaining the distinction and grayscale range between the background area and the foreground area, so that a significant contrast improvement can be achieved between the foreground area and the background area.

[0116] Additionally, in one embodiment, reference is made to Figure 5 ,exist Figure 1 In step S4 of the illustrated embodiment, the first transmittance map, the second transmittance map, and the third transmittance map are merged to obtain a contrast-enhanced image of the remote sensing image. The method also includes the following steps:

[0117] S41: Obtain the degraded image of the remote sensing image and obtain the atmospheric light component of the degraded image;

[0118] S42: Obtain a contrast-enhanced image based on the degraded image, atmospheric light component, first transmittance map, second transmittance map, and third transmittance map.

[0119] It should be noted that the contrast-enhanced image is obtained from the degraded image, atmospheric light component, first transmittance map, second transmittance map, and third transmittance map, and is expressed by the following ninth formula:

[0120] ;

[0121] in, To enhance the contrast of the image, For degraded images, The first, second, and third transmittance maps are stitched together. For fuzzy Gaussian kernel, For the atmospheric light component, the solution was obtained in the previous section. Medium parameters The settings ensure that the magnitude of the atmospheric light component does not affect the image enhancement effect.

[0122] Those skilled in the art will understand that all or some of the steps and systems in the methods disclosed above can be implemented as software, firmware, hardware, and suitable combinations thereof. Some or all of the physical components can be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit. Such software can be distributed on a computer-readable medium, which can include computer storage media (or non-transitory media) and communication media (or transient media). As is known to those skilled in the art, the term computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disc (DVD) or other optical disc storage, magnetic cartridges, magnetic tape, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and is accessible to a computer. Furthermore, as is known to those skilled in the art, communication media typically include computer-readable instructions, data structures, program modules, or other data in modulated data signals such as carrier waves or other transmission mechanisms, and may include any information delivery medium.

[0123] The systems, apparatuses, modules, or units described in one or more of the above embodiments may be implemented by a computer chip or entity, or by a product having a certain function. A typical implementation device is a computer. Specifically, a computer may be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smartphone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or any combination of these devices.

[0124] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0125] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0126] In summary, the above description is merely a preferred embodiment of this specification and is not intended to limit the scope of protection of this specification. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this specification should be included within the scope of protection of this specification.

Claims

1. A method for enhancing the contrast of remotely sensed images, characterized in that, include: S1: Obtain the first grayscale threshold of the remote sensing image, and divide the remote sensing image into saturated regions and under-enhanced regions according to the first grayscale threshold; S2: Obtain the pixel grayscale value of the remote sensing image, and divide the under-enhanced region into a foreground region and a background region according to the pixel grayscale value. The foreground region is provided with a first grayscale range, and the background region is provided with a second grayscale range. The lower limit of the first grayscale range is higher than the upper limit of the second grayscale range. S3: Construct a first optimization model and configure a first transmittance map of the foreground region using the first optimization model. The first optimization model is used to enhance the pixel grayscale values ​​of the first transmittance map within the first grayscale range. Obtain a second grayscale threshold and the average pixel grayscale value of the background region. Configure a second transmittance map of the background region based on the second grayscale threshold and the average pixel grayscale value. Reduce the pixel grayscale values ​​of the second transmittance map within the second grayscale range. Set the transmittance of the saturated region to 1 to obtain a third transmittance map of the saturated region. S4: The third transmittance map is merged with the adjusted first transmittance map and the second transmittance map to obtain the contrast-enhanced image of the remote sensing image.

2. The method for enhancing the contrast of remote sensing images according to claim 1, characterized in that, The step of constructing a first optimization model and configuring a first transmittance map of the foreground region using the first optimization model includes: S31: Obtain the degraded image of the remote sensing image, and determine the bright channel value, dark channel value and auxiliary parameters of the degraded image; S32: Calculate the lower limit of the transmittance of the foreground region based on the bright channel value, the dark channel value, and the auxiliary parameters; S33: Construct a mask matrix based on a preset Gaussian kernel and the degraded image; S34: Obtain the gradient of the first transmittance map, construct the first optimization model based on the gradient of the first transmittance map, the mask matrix, the lower limit of transmittance, and the preset first regularization parameter, second regularization parameter, and upper limit of transmittance, solve the foreground region using the first optimization model to obtain the optimized transmittance vector of the foreground region, and obtain the first transmittance map based on the optimized transmittance vector.

3. The method for enhancing the contrast of remote sensing images according to claim 2, characterized in that, The lower limit of transmittance of the foreground region is calculated based on the bright channel value, the dark channel value, and the auxiliary parameters, and is expressed by the following formula: ； in, Lower limit of transmittance For the set of neighboring pixels, For pixels that are within the neighboring pixel set, To highlight channel values, For dark channel value, These are auxiliary parameters.

4. The method for enhancing the contrast of remote sensing images according to claim 2, characterized in that, The construction of the first optimization model is expressed by the following formula: ； in, For the first regularization parameter, For the second regularization parameter, For the first transmittance diagram, Lower limit of transmittance The preset upper limit of transmittance, For mask matrix, The gradient of the first transmittance map, To smooth out constraint terms, The norm regularization parameter.

5. The method for enhancing the contrast of remote sensing images according to claim 2, characterized in that, The first optimization model solves for the foreground region to obtain the optimized transmittance vector of the foreground region, including: S341: Convert the first optimization model into a second optimization model, and solve the foreground region according to the second optimization model to obtain the gradient transmittance map of the foreground region; S342: Convert the mask matrix into a sparse matrix; convert the upper limit of transmittance into an upper limit vector of transmittance; convert the lower limit of transmittance into a lower limit vector of transmittance; convert the gradient transmittance map into a first gradient transmittance vector and a second gradient transmittance vector; S343: Obtain the preset first gradient sparse operator and second gradient sparse operator, and obtain the optimized transmittance vector based on the sparse matrix, the transmittance upper limit vector, the transmittance lower limit vector, the first gradient transmittance vector, the second gradient transmittance vector, the first gradient sparse operator, and the second gradient sparse operator.

6. The method for enhancing the contrast of remote sensing images according to claim 1, characterized in that, The step of configuring the second transmittance map of the background region based on the second grayscale threshold and the average pixel grayscale value includes: S35: Obtain all the pixel grayscale values ​​within the background area, and obtain the average pixel grayscale value of the image block based on the pixel grayscale values; S36: Correct the difference between the median gray level and the mean gray level of the pixels according to preset empirical parameters, so as to adjust the pixel gray level of the image block of the background region and obtain the second transmittance map of the background region.

7. The method for enhancing the contrast of remote sensing images according to claim 6, characterized in that, The correction of the difference between the median gray level and the mean gray level of pixels based on preset empirical parameters is expressed by the following formula: ； in, For the second transmittance map, For empirical parameters, The median of grayscale This represents the average grayscale value of the pixels.

8. The method for enhancing the contrast of remote sensing images according to claim 1, characterized in that, The step of merging the third transmittance map with the adjusted first and second transmittance maps to obtain the contrast-enhanced image of the remote sensing image includes: S41: Obtain the degraded image of the remote sensing image, and obtain the atmospheric light component of the degraded image; S42: Obtain the contrast-enhanced image based on the degraded image, the atmospheric light component, the first transmittance map, the second transmittance map, and the third transmittance map.

Images