A display device system and a method for adaptively enhancing image quality

Abstract

This disclosure discloses a display device system and an adaptive image quality enhancement method for maximizing the image quality enhancement of different images. The method includes: acquiring an original image; performing iterative detail enhancement processing on the original image using at least one filtering function containing iteratively updated filtering parameters; when the difference in quality evaluation scores between the input image and the detail-enhanced image of the i-th iteration satisfies an iteration condition, updating the filtering parameters of the i-th iteration to obtain the filtering parameters of the (i+1)-th iteration; and continuing to perform detail enhancement processing on the input image using at least one filtering function containing the filtering parameters of the (i+1)-th iteration until the difference in quality evaluation scores between the input image and the detail-enhanced image of the (i+1)-th iteration satisfies the iteration condition; the display device system further includes a display screen configured to display the detail-enhanced image of the last iteration.

Description

Technical Field

[0001] This disclosure relates to the field of image processing technology, and in particular to a display device system and a method for adaptively enhancing image quality. Background Technology

[0002] Image enhancement works by adjusting the contrast, saturation, and color of an image to improve its quality and visual clarity. Currently, image enhancement solutions used in display devices require continuous tuning and parameter adjustments based on different image types and content. When using such a solution, the adjusted parameters are applied to enhance the quality of the input image.

[0003] Current image enhancement solutions use fixed parameters for any input image. However, different images have different contrast, saturation, and color. If the same set of parameters is used to enhance different images, the enhancement effect will be poor. It is impossible to adaptively adjust the parameters according to different input images and perform image enhancement. Summary of the Invention

[0004] This disclosure provides a display device system and an adaptive image quality enhancement method, which performs adaptive image quality enhancement processing based on the edge features and high-frequency information of different input images, thereby maximizing the image quality enhancement of different images.

[0005] In a first aspect, an embodiment of this disclosure provides a display device, comprising:

[0006] Memory on which computer-readable instructions are stored;

[0007] When a processor is configured to execute the computer-readable instructions, it performs the following steps:

[0008] Acquire the original image, and perform iterative detail enhancement processing on the original image using at least one filtering function containing iteratively updated filtering parameters, specifically including:

[0009] For the input image of the detail enhancement processing, at least one filtering function containing the filtering parameters updated in the i-th iteration is used to extract the feature information of the input image in the i-th iteration, and the feature information of the i-th iteration is used to perform the detail enhancement processing on the input image in the i-th iteration to obtain the detail-enhanced image in the i-th iteration; wherein, the input image includes an image related to the original image, and i is an integer greater than or equal to 1;

[0010] When the difference between the quality evaluation scores of the input image and the detail-enhanced image of the i-th iteration satisfies the iteration condition, the filtering parameters of the i-th iteration are updated to obtain the filtering parameters of the (i+1)-th iteration. The input image is then further enhanced with at least one filtering function containing the filtering parameters of the (i+1)-th iteration until the difference between the quality evaluation scores of the input image and the detail-enhanced image of the (i+1)-th iteration satisfies the iteration condition.

[0011] The display device system also includes a display screen configured to display a detail-enhanced image of the last iteration.

[0012] After acquiring the original image, before performing iterative detail enhancement processing on the original image using at least one filtering function containing iteratively updated filtering parameters, the processor is specifically configured to execute:

[0013] When the original image is determined to be an RGB image, the original RGB image is converted to HSV color channels to obtain the original HSV image;

[0014] The brightness V channel image of the original HSV image is used as the input image for the detail enhancement process.

[0015] As one possible implementation, before displaying the image from the last iteration of detail enhancement processing, the processor is specifically further configured to perform:

[0016] Convert the image processed in the last iteration of detail enhancement to an RGB detail-enhanced image;

[0017] The display screen is specifically configured to perform:

[0018] Display the RGB detail-enhanced image.

[0019] As one possible implementation, the processor is specifically configured to execute:

[0020] Using a first filtering function that includes the first filtering parameters updated in the i-th iteration, the first feature information of the original image in the i-th iteration is extracted. The first feature information of the i-th iteration is then used to perform detail enhancement processing on the original image in the i-th iteration to obtain the first enhanced image in the i-th iteration.

[0021] Using a second filtering function that includes the second filtering parameters updated in the i-th iteration, the second feature information of the first enhanced image in the i-th iteration is extracted. The first enhanced image is then subjected to detail enhancement processing in the i-th iteration using the second feature information of the i-th iteration, resulting in the detail-enhanced image in the i-th iteration.

[0022] As one possible implementation, the processor is specifically configured to execute:

[0023] The first feature information of the i-th iteration is superimposed on the original image to obtain the first enhanced image of the i-th iteration;

[0024] The second feature information of the i-th iteration is superimposed on the first enhanced image of the i-th iteration to obtain the detail enhanced image of the i-th iteration.

[0025] As one possible implementation, the controller is specifically configured to perform:

[0026] The original image is filtered using a first filtering function that includes the first filtering parameters updated in the i-th iteration to obtain the first filtered image in the i-th iteration; the original image is subtracted from the first filtered image in the i-th iteration to obtain the first feature information in the i-th iteration.

[0027] The first enhanced image of the i-th iteration is filtered using a second filtering function that includes the second filtering parameters updated in the i-th iteration to obtain the second filtered image of the i-th iteration; the first enhanced image of the i-th iteration is subtracted from the second filtered image of the i-th iteration to obtain the second feature information of the i-th iteration.

[0028] As one possible implementation, the first filtering function is a guided filtering function, and the second filtering function is a Gaussian filtering function; or,

[0029] The first filtering function is a Gaussian filtering function, and the second filtering function is a guided filtering function.

[0030] As one possible implementation, the quality assessment score of the original image is a first NIQE score obtained by evaluating the quality of the original image using the No Reference Image Quality Assessment (NIQE) algorithm.

[0031] The quality evaluation score of the detail-enhanced image in the i-th iteration is the second NIQE score obtained by evaluating the quality of the detail-enhanced image in the i-th iteration using the NIQE algorithm.

[0032] As one possible implementation, the controller is specifically configured to perform:

[0033] The difference between the first NIQE score and the second NIQE score is determined to be less than or equal to zero or greater than or equal to a preset threshold.

[0034] As one possible implementation, after acquiring the original image and before performing iterative detail enhancement processing on the original image using at least one filtering function containing iteratively updated filtering parameters, the controller is specifically configured to perform:

[0035] Using a super-resolution algorithm, the resolution of the input image for super-resolution processing is reconstructed to a preset resolution to obtain a reconstructed image, wherein the preset resolution is higher than the resolution of the original image; wherein the input image for super-resolution processing is either the original image or a contrast-enhanced image, and the contrast-enhanced image is obtained by performing contrast enhancement processing on the original image;

[0036] The reconstructed image is used as the input image for the detail enhancement process.

[0037] As one possible implementation, the controller is specifically configured to perform:

[0038] The super-resolution factor corresponding to the size is determined based on the size of the original image;

[0039] Based on the super-resolution factor and the super-resolution algorithm, the resolution of the original image is reconstructed to a preset resolution.

[0040] As one possible implementation, after acquiring the original image and before performing iterative detail enhancement processing on the original image using at least one filtering function containing iteratively updated filtering parameters, the controller is specifically configured to perform:

[0041] The reconstructed image after super-resolution processing or the original image without super-resolution processing is used as the input image, and iterative contrast enhancement processing is performed using the iteratively updated gamma value:

[0042] The input image for the contrast enhancement process is subjected to gamma transformation according to the gamma value updated in the j-th iteration to obtain the contrast-enhanced image in the j-th iteration; where j is an integer greater than or equal to 1.

[0043] If the contrast difference between the contrast-enhanced image of the j-th iteration and the contrast-enhanced input image is greater than the contrast threshold, the contrast of the contrast-enhanced input image is replaced with the contrast of the contrast-enhanced image of the j-th iteration, and the gamma value of the j-th iteration is updated to obtain the gamma value of the (j+1)-th iteration. Using the gamma value of the (j+1)-th iteration, the contrast-enhanced input image is further subjected to iterative contrast enhancement processing until the contrast difference between the contrast-enhanced image of the j-th iteration and the contrast-enhanced input image is less than or equal to the contrast threshold.

[0044] The contrast-enhanced image from the last iteration is used as the input image for the detail enhancement process.

[0045] As one possible implementation, before using the contrast-enhanced image from the last iteration as the input image for the detail enhancement process, the controller is specifically further configured to perform:

[0046] Determine the difference in the mean brightness between the input image of the contrast enhancement process and the contrast enhancement image of the last iteration;

[0047] The difference is superimposed with the brightness of the contrast-enhanced image from the last iteration to obtain a superimposed image;

[0048] The overlaid image is used as the input image for the detail enhancement process.

[0049] As one possible implementation, prior to the iterative contrast enhancement processing using iteratively updated gamma values, the controller is further configured to perform:

[0050] When the reconstructed image after super-resolution processing is determined to be an RGB image, the reconstructed image is converted to HSV to obtain the original HSV image, and the V channel image of the original HSV image is used as the input image for the contrast enhancement processing; or,

[0051] When the original image that has not undergone super-resolution processing is determined to be an RGB image, the original image is converted to HSV to obtain an original HSV image, and the V channel image of the original HSV image is used as the input image for the contrast enhancement processing.

[0052] As one possible implementation, before using the overlaid image as the input image for the detail enhancement process, the controller is specifically configured to perform:

[0053] The overlay image is converted into an RGB overlay image.

[0054] Secondly, this disclosure provides an adaptive image quality enhancement method, the method comprising:

[0055] The process involves acquiring an original image and performing iterative detail enhancement processing on the original image using at least one filtering function containing iteratively updated filtering parameters. Specifically, this includes: extracting feature information of the input image for the detail enhancement processing using at least one filtering function containing filtering parameters updated in the i-th iteration; performing detail enhancement processing on the input image for the i-th iteration using the feature information from the i-th iteration to obtain a detail-enhanced image for the i-th iteration; wherein the input image includes images related to the original image, and i is an integer greater than or equal to 1.

[0056] When the difference between the quality evaluation scores of the input image and the detail-enhanced image of the i-th iteration satisfies the iteration condition, the filtering parameters of the i-th iteration are updated to obtain the filtering parameters of the (i+1)-th iteration. The input image is then further enhanced with at least one filtering function containing the filtering parameters of the (i+1)-th iteration until the difference between the quality evaluation scores of the input image and the detail-enhanced image of the (i+1)-th iteration satisfies the iteration condition.

[0057] Shows an enhanced image of the last iteration.

[0058] As one possible implementation, after acquiring the original image and before performing iterative detail enhancement processing on the original image using at least one filtering function containing iteratively updated filtering parameters, the method further includes:

[0059] When the original image is determined to be an RGB image, the original RGB image is converted to HSV color channels to obtain the original HSV image;

[0060] The brightness V channel image of the original HSV image is used as the input image for the detail enhancement process.

[0061] As one possible implementation, before displaying the image from the last iteration of detail enhancement processing, the method further includes:

[0062] Convert the image processed in the last iteration of detail enhancement to an RGB detail-enhanced image;

[0063] The detailed enhanced image showing the last iteration specifically includes:

[0064] Display the RGB detail-enhanced image.

[0065] As one possible implementation, the step of extracting feature information of the original image in the i-th iteration using at least one filtering function containing the filtering parameters updated in the i-th iteration, and then using the feature information of the i-th iteration to perform detail enhancement processing on the original image in the i-th iteration to obtain a detail-enhanced image in the i-th iteration specifically includes:

[0066] Using a first filtering function that includes the first filtering parameters updated in the i-th iteration, the first feature information of the original image in the i-th iteration is extracted. The first feature information of the i-th iteration is then used to perform detail enhancement processing on the original image in the i-th iteration to obtain the first enhanced image in the i-th iteration.

[0067] Using a second filtering function that includes the second filtering parameters updated in the i-th iteration, the second feature information of the first enhanced image in the i-th iteration is extracted. The first enhanced image is then subjected to detail enhancement processing in the i-th iteration using the second feature information of the i-th iteration, resulting in the detail-enhanced image in the i-th iteration.

[0068] As one possible implementation, the original image is subjected to detail enhancement processing using the first feature information from the i-th iteration to obtain the first enhanced image of the i-th iteration, including:

[0069] The first feature information of the i-th iteration is superimposed on the original image to obtain the first enhanced image of the i-th iteration;

[0070] The step of performing detail enhancement processing on the first enhanced image using the second feature information from the i-th iteration includes:

[0071] The second feature information of the i-th iteration is superimposed on the first enhanced image of the i-th iteration to obtain the detail enhanced image of the i-th iteration.

[0072] As one possible implementation, extracting the first feature information of the original image in the i-th iteration using a first filtering function that includes the first filtering parameters updated in the i-th iteration includes:

[0073] The original image is filtered using a first filtering function that includes the first filtering parameters updated in the i-th iteration to obtain the first filtered image in the i-th iteration; the original image is subtracted from the first filtered image in the i-th iteration to obtain the first feature information in the i-th iteration.

[0074] The step of extracting the second feature information of the first enhanced image in the i-th iteration using a second filtering function that includes the second filtering parameters updated in the i-th iteration includes:

[0075] The first enhanced image of the i-th iteration is filtered using a second filtering function that includes the second filtering parameters updated in the i-th iteration to obtain the second filtered image of the i-th iteration; the first enhanced image of the i-th iteration is subtracted from the second filtered image of the i-th iteration to obtain the second feature information of the i-th iteration.

[0076] As one possible implementation, the first filtering function is a guided filtering function, and the second filtering function is a Gaussian filtering function; or,

[0077] The first filtering function is a Gaussian filtering function, and the second filtering function is a guided filtering function.

[0078] As one possible implementation, the quality assessment score of the original image is a first NIQE score obtained by evaluating the quality of the original image using the No Reference Image Quality Assessment (NIQE) algorithm.

[0079] The quality evaluation score of the detail-enhanced image in the i-th iteration is the second NIQE score obtained by evaluating the quality of the detail-enhanced image in the i-th iteration using the NIQE algorithm.

[0080] As one possible implementation, determining that the difference between the quality evaluation scores of the original image and the detail-enhanced image of the i-th iteration satisfies the iteration condition includes:

[0081] The difference between the first NIQE score and the second NIQE score is determined to be less than or equal to zero or greater than or equal to a preset threshold.

[0082] As one possible implementation, after acquiring the original image and before performing iterative detail enhancement processing on the original image using at least one filtering function containing iteratively updated filtering parameters, the method further includes:

[0083] Using a super-resolution algorithm, the resolution of the input image for super-resolution processing is reconstructed to a preset resolution to obtain a reconstructed image, wherein the preset resolution is higher than the resolution of the original image; wherein the input image for super-resolution processing is either the original image or a contrast-enhanced image, and the contrast-enhanced image is obtained by performing contrast enhancement processing on the original image;

[0084] The reconstructed image is used as the input image for the detail enhancement process.

[0085] As one possible implementation, the step of using a super-resolution algorithm to reconstruct the resolution of the original image to a preset resolution includes:

[0086] The super-resolution factor corresponding to the size is determined based on the size of the original image;

[0087] Based on the super-resolution factor and the super-resolution algorithm, the resolution of the original image is reconstructed to a preset resolution.

[0088] As one possible implementation, after acquiring the original image and before performing iterative detail enhancement processing on the original image using at least one filtering function containing iteratively updated filtering parameters, the method further includes:

[0089] The reconstructed image after super-resolution processing or the original image without super-resolution processing is used as the input image, and iterative contrast enhancement processing is performed using the iteratively updated gamma value:

[0090] The input image for the contrast enhancement process is subjected to gamma transformation according to the gamma value updated in the j-th iteration to obtain the contrast-enhanced image in the j-th iteration; where j is an integer greater than or equal to 1.

[0091] If the contrast difference between the contrast-enhanced image of the j-th iteration and the contrast-enhanced input image is greater than the contrast threshold, the contrast of the contrast-enhanced input image is replaced with the contrast of the contrast-enhanced image of the j-th iteration, and the gamma value of the j-th iteration is updated to obtain the gamma value of the (j+1)-th iteration. Using the gamma value of the (j+1)-th iteration, the contrast-enhanced input image is further subjected to iterative contrast enhancement processing until the contrast difference between the contrast-enhanced image of the j-th iteration and the contrast-enhanced input image is less than or equal to the contrast threshold.

[0092] The contrast-enhanced image from the last iteration is used as the input image for the detail enhancement process.

[0093] As one possible implementation, before using the contrast-enhanced image from the last iteration as the input image for the detail enhancement process, the method further includes:

[0094] Determine the difference in the mean brightness between the input image of the contrast enhancement process and the contrast enhancement image of the last iteration;

[0095] The difference is superimposed with the brightness of the contrast-enhanced image from the last iteration to obtain a superimposed image;

[0096] The overlaid image is used as the input image for the detail enhancement process.

[0097] As one possible implementation, prior to the iterative contrast enhancement processing using iteratively updated gamma values, the method further includes:

[0098] When the reconstructed image after super-resolution processing is determined to be an RGB image, the reconstructed image is converted to HSV to obtain the original HSV image, and the V channel image of the original HSV image is used as the input image for the contrast enhancement processing; or,

[0099] When the original image that has not undergone super-resolution processing is determined to be an RGB image, the original image is converted to HSV to obtain an original HSV image, and the V channel image of the original HSV image is used as the input image for the contrast enhancement processing.

[0100] As one possible implementation, before using the overlaid image as the input image for the detail enhancement process, the method further includes:

[0101] The overlay image is converted into an RGB overlay image.

[0102] Thirdly, embodiments of this disclosure also provide an adaptive image quality enhancement device, comprising:

[0103] An enhancement processing unit is configured to acquire an original image and perform iterative detail enhancement processing on the original image using at least one filtering function containing iteratively updated filtering parameters. Specifically, this includes: extracting feature information of the input image for the detail enhancement processing using at least one filtering function containing filtering parameters updated in the i-th iteration, and performing detail enhancement processing on the input image for the i-th iteration using the feature information of the i-th iteration to obtain a detail-enhanced image for the i-th iteration; wherein the input image includes an image related to the original image, and i is an integer greater than or equal to 1.

[0104] When the difference between the quality evaluation scores of the input image and the detail-enhanced image of the i-th iteration satisfies the iteration condition, the filtering parameters of the i-th iteration are updated to obtain the filtering parameters of the (i+1)-th iteration. The input image is then further enhanced with at least one filtering function containing the filtering parameters of the (i+1)-th iteration until the difference between the quality evaluation scores of the input image and the detail-enhanced image of the (i+1)-th iteration satisfies the iteration condition.

[0105] Enhanced output unit for displaying a detailed, enhanced image from the last iteration.

[0106] As one possible implementation, before displaying the image from the last iteration of detail enhancement processing, a transformation unit is further included specifically for:

[0107] Convert the image processed in the last iteration of detail enhancement to an RGB detail-enhanced image;

[0108] The enhanced output unit is further specifically used for:

[0109] Display the RGB detail-enhanced image.

[0110] As one possible implementation, the enhancement processing unit is specifically used for:

[0111] Using a first filtering function that includes the first filtering parameters updated in the i-th iteration, the first feature information of the original image in the i-th iteration is extracted. The first feature information of the i-th iteration is then used to perform detail enhancement processing on the original image in the i-th iteration to obtain the first enhanced image in the i-th iteration.

[0112] Using a second filtering function that includes the second filtering parameters updated in the i-th iteration, the second feature information of the first enhanced image in the i-th iteration is extracted. The first enhanced image is then subjected to detail enhancement processing in the i-th iteration using the second feature information of the i-th iteration, resulting in the detail-enhanced image in the i-th iteration.

[0113] As one possible implementation, the enhancement processing unit is specifically used for:

[0114] The first feature information of the i-th iteration is superimposed on the original image to obtain the first enhanced image of the i-th iteration;

[0115] The second feature information of the i-th iteration is superimposed on the first enhanced image of the i-th iteration to obtain the detail enhanced image of the i-th iteration.

[0116] As one possible implementation, the enhancement processing unit is specifically used for:

[0117] The original image is filtered using a first filtering function that includes the first filtering parameters updated in the i-th iteration to obtain the first filtered image in the i-th iteration; the original image is subtracted from the first filtered image in the i-th iteration to obtain the first feature information in the i-th iteration.

[0118] The first enhanced image of the i-th iteration is filtered using a second filtering function that includes the second filtering parameters updated in the i-th iteration to obtain the second filtered image of the i-th iteration; the first enhanced image of the i-th iteration is subtracted from the second filtered image of the i-th iteration to obtain the second feature information of the i-th iteration.

[0119] As one possible implementation, the first filtering function is a guided filtering function, and the second filtering function is a Gaussian filtering function; or,

[0120] The first filtering function is a Gaussian filtering function, and the second filtering function is a guided filtering function.

[0121] As one possible implementation, the quality assessment score of the original image is a first NIQE score obtained by evaluating the quality of the original image using the No Reference Image Quality Assessment (NIQE) algorithm.

[0122] The quality evaluation score of the detail-enhanced image in the i-th iteration is the second NIQE score obtained by evaluating the quality of the detail-enhanced image in the i-th iteration using the NIQE algorithm.

[0123] As one possible implementation, the difference determination unit 901 is specifically used for:

[0124] The difference between the first NIQE score and the second NIQE score is determined to be less than or equal to zero or greater than or equal to a preset threshold.

[0125] As one possible implementation, after acquiring the original image and before performing iterative detail enhancement processing on the original image using at least one filtering function containing iteratively updated filtering parameters, a reconstruction unit is further included, specifically for:

[0126] Using a super-resolution algorithm, the resolution of the input image for super-resolution processing is reconstructed to a preset resolution to obtain a reconstructed image, wherein the preset resolution is higher than the resolution of the original image; wherein the input image for super-resolution processing is either the original image or a contrast-enhanced image, and the contrast-enhanced image is obtained by performing contrast enhancement processing on the original image;

[0127] The reconstructed image is used as the input image for the detail enhancement process.

[0128] As one possible implementation, the reconstruction unit is specifically used for:

[0129] The super-resolution factor corresponding to the size is determined based on the size of the original image;

[0130] Based on the super-resolution factor and the super-resolution algorithm, the resolution of the original image is reconstructed to a preset resolution.

[0131] As one possible implementation, after acquiring the original image and before performing iterative detail enhancement processing on the original image using at least one filtering function containing iteratively updated filtering parameters, a contrast enhancement unit is further included, specifically for:

[0132] The reconstructed image after super-resolution processing or the original image without super-resolution processing is used as the input image, and iterative contrast enhancement processing is performed using the iteratively updated gamma value:

[0133] The input image for the contrast enhancement process is subjected to gamma transformation according to the gamma value updated in the j-th iteration to obtain the contrast-enhanced image in the j-th iteration; where j is an integer greater than or equal to 1.

[0134] If the contrast difference between the contrast-enhanced image of the j-th iteration and the contrast-enhanced input image is greater than the contrast threshold, the contrast of the contrast-enhanced input image is replaced with the contrast of the contrast-enhanced image of the j-th iteration, and the gamma value of the j-th iteration is updated to obtain the gamma value of the (j+1)-th iteration. Using the gamma value of the (j+1)-th iteration, the contrast-enhanced input image is further subjected to iterative contrast enhancement processing until the contrast difference between the contrast-enhanced image of the j-th iteration and the contrast-enhanced input image is less than or equal to the contrast threshold.

[0135] The contrast-enhanced image from the last iteration is used as the input image for the detail enhancement process.

[0136] As one possible implementation, before using the contrast-enhanced image from the last iteration as the input image for the detail enhancement process, the contrast enhancement unit is further configured to:

[0137] Determine the difference in the mean brightness between the input image of the contrast enhancement process and the contrast enhancement image of the last iteration;

[0138] The difference is superimposed with the brightness of the contrast-enhanced image from the last iteration to obtain a superimposed image;

[0139] The overlaid image is used as the input image for the detail enhancement process.

[0140] As one possible implementation, before the iterative contrast enhancement processing using the iteratively updated gamma value, the contrast enhancement unit is further configured to:

[0141] When the reconstructed image after super-resolution processing is determined to be an RGB image, the reconstructed image is converted to HSV to obtain the original HSV image, and the V channel image of the original HSV image is used as the input image for the contrast enhancement processing; or,

[0142] When the original image that has not undergone super-resolution processing is determined to be an RGB image, the original image is converted to HSV to obtain an original HSV image, and the V channel image of the original HSV image is used as the input image for the contrast enhancement processing.

[0143] As one possible implementation, before using the overlaid image as the input image for the detail enhancement process, the contrast enhancement unit is further configured to:

[0144] The overlay image is converted into an RGB overlay image.

[0145] Fourthly, embodiments of this disclosure also provide a non-transient computer storage medium having a computer program stored thereon, which, when executed by a processor, is used to implement the steps of the method described in the first aspect above.

[0146] These or other aspects of this disclosure will become more apparent in the following description of embodiments. Attached Figure Description

[0147] To more clearly illustrate the technical solutions in the embodiments of this disclosure, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this disclosure. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0148] Figure 1 A schematic diagram of a display device provided in an embodiment of this disclosure;

[0149] Figure 2 An iterative flowchart of an adaptive enhancement process provided in an embodiment of this disclosure;

[0150] Figure 3 A flowchart illustrating the implementation of enhancing the V channel image of the original image according to an embodiment of this disclosure;

[0151] Figure 4 This disclosure provides a flowchart illustrating a specific implementation of a single enhancement process.

[0152] Figure 5 A flowchart illustrating the implementation of a super-resolution reconstruction method provided in this disclosure embodiment;

[0153] Figure 6 A flowchart illustrating an implementation method for contrast enhancement provided in this disclosure.

[0154] Figure 7 A flowchart illustrating a specific implementation of three enhancement processing methods provided in this disclosure embodiment;

[0155] Figure 8 A flowchart illustrating an implementation of an adaptive image quality enhancement method provided in this disclosure.

[0156] Figure 9This is a schematic diagram of an adaptive image quality enhancement device provided in an embodiment of this disclosure. Detailed Implementation

[0157] To make the objectives, technical solutions, and advantages of this disclosure clearer, the disclosure will be further described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this disclosure, and not all of them. Based on the embodiments of this disclosure, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this disclosure.

[0158] In this disclosure, the term "and / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent three cases: A alone, A and B simultaneously, and B alone. The character " / " generally indicates that the preceding and following related objects have an "or" relationship.

[0159] The application scenarios described in this disclosure are for the purpose of more clearly illustrating the technical solutions of this disclosure and do not constitute a limitation on the technical solutions provided in this disclosure. Those skilled in the art will understand that with the emergence of new application scenarios, the technical solutions provided in this disclosure are also applicable to similar technical problems. In the description of this disclosure, unless otherwise stated, "multiple" means two or more.

[0160] Image enhancement works by adjusting the contrast, saturation, and color of an image to improve its visual clarity. Currently, image enhancement solutions used in display devices require continuous tuning and parameter adjustments based on different image types and content. When using such a solution, the adjusted parameters are applied to the input image. However, once the algorithm is delivered to the display device, the corresponding adjusted parameters remain fixed. Currently, image enhancement is typically used when display devices offer users an image enhancement option. Since the parameters of these algorithms are fixed, after receiving a user's image enhancement command, applying the same fixed-parameter algorithm to images with different edge features, high-frequency information, resolutions, contrasts, and sizes can easily lead to poor image enhancement results, a poor user experience, and hinder subsequent algorithm optimization. Current image enhancement solutions use fixed parameters for any input image. However, different images differ in contrast, saturation, color, and other parameters. Using the same set of parameters to enhance different images can easily result in poor image enhancement effects, as it cannot adaptively adjust parameters based on different input images.

[0161] The adaptive image quality enhancement method provided in this embodiment can more specifically enhance the image quality of the input image based on the feature information of different images, effectively improving the adaptive capability of the image quality enhancement algorithm. The core algorithm of this embodiment utilizes an iterative enhancement approach. It performs iterative enhancement based on whether the quality evaluation scores of the original image and the enhanced image are close. In each iteration, the original image is enhanced using its current feature information. Then, based on whether the quality evaluation scores of the original image and the enhanced image are close, it is determined whether to continue enhancing the original image. If enhancement continues, the current feature information is updated, and the enhanced image is further enhanced using the updated feature information until the quality evaluation scores of the original image and the enhanced image are close, at which point the iteration stops. The image after the last enhancement is output as the enhanced image of the input original image. In this embodiment, "current iteration" specifically refers to the i-th iteration in the iterative detail enhancement process, where i ≥ 1.

[0162] In this embodiment, each enhancement process requires updating various filter parameters. These updated parameters are then used to update various feature information, which is then used to further enhance the original image. The image quality enhancement algorithm in this embodiment can enhance the original image based on its own feature information, using extracted feature information to further enhance the original image, thereby making the original image visually clearer and achieving the goal of image quality enhancement. Because it can enhance some semantic information (feature information) of the original image itself, it can maximize the image quality enhancement of the input original image, effectively improving the image quality enhancement effect.

[0163] The adaptive image quality enhancement method provided in this embodiment is mainly applied to display devices, including but not limited to large-size (generally 50 inches or more) intelligent interactive display devices, large-size electronic whiteboards, and large-size displays (which can be splicing screens).

[0164] like Figure 1 As shown in the figure, a display device system provided in this disclosure includes a memory 100, a processor 101, and a display screen 102, wherein:

[0165] Memory 100, on which computer-readable instructions are stored;

[0166] When processor 101 is configured to execute the computer-readable instructions, it performs the following steps:

[0167] Acquire the original image, and perform iterative detail enhancement processing on the original image using at least one filtering function containing iteratively updated filtering parameters, specifically including:

[0168] For the input image of the detail enhancement processing, at least one filtering function containing the filtering parameters updated in the i-th iteration is used to extract the feature information of the input image in the i-th iteration, and the feature information of the i-th iteration is used to perform the detail enhancement processing on the input image in the i-th iteration to obtain the detail-enhanced image in the i-th iteration; wherein, the input image includes an image related to the original image, and i is an integer greater than or equal to 1;

[0169] When the difference between the quality evaluation scores of the input image and the detail-enhanced image of the i-th iteration satisfies the iteration condition, the filtering parameters of the i-th iteration are updated to obtain the filtering parameters of the (i+1)-th iteration. The input image is then further enhanced with at least one filtering function containing the filtering parameters of the (i+1)-th iteration until the difference between the quality evaluation scores of the input image and the detail-enhanced image of the (i+1)-th iteration satisfies the iteration condition.

[0170] Display screen 102 is configured to display a detailed enhanced image of the last iteration.

[0171] In some embodiments, the input image for detail enhancement processing is an image related to the original image, including but not limited to the original image, the V channel image of the original HSV image, etc.

[0172] It should be noted that in this embodiment, the process of image enhancement by the display device is imperceptible to the user. The display screen 102 can directly display the image after the detail enhancement process is completed, thereby achieving image enhancement processing.

[0173] In implementation, the feature information and the filtering parameters are uniquely matched, and the filtering function and the filtering parameters are also uniquely matched. The filtering parameters of at least one initial filtering function are preset. In this embodiment, the filtering parameters are used to represent the size of the filtering window. For any filtering parameter in this embodiment, the way to update the filtering parameters in this example is to gradually increase the size of the filtering window by superimposing a preset step size on the filtering parameters of the i-th iteration; or to gradually decrease the size of the filtering window by subtracting the preset step size from the filtering parameters of the i-th iteration. This embodiment does not impose too many restrictions on the specific way of updating the filtering parameters.

[0174] In some embodiments, feature information in the original image is extracted based on the filtering parameters of at least one filtering function, thereby determining the feature information of the original image. The feature information in this embodiment includes, but is not limited to, at least one of edge feature information and high-frequency information.

[0175] In some embodiments, the filtering function in this embodiment includes, but is not limited to, at least one of: a guided filtering function and a Gaussian filtering function.

[0176] In some embodiments, the Natural Image Quality Evaluator (NIQE) algorithm is used to evaluate image quality. NIQE is a no-reference image quality assessment method that calculates the spatial NSS (Natural Scene Statistics) features of a natural image, then fits a multivariate Gaussian model (MVG). Using 125 natural images, the mean and covariance of the MVG are calculated. This MVG is then used to determine the NIQE score of the input image. The weights in this MVG are publicly available. When calculating the NIQE score, the mean and variance of the MVG for the test image (including the original image and the image after detail enhancement) are first calculated. Then, the distance between the test image and the natural image is calculated. A closer distance corresponds to a lower NIQE score, indicating a more realistic and higher-quality test image.

[0177] In this embodiment, the quality assessment score of the original image is the first NIQE score obtained by evaluating the quality of the original image using the No Reference Image Quality Assessment (NIQE) algorithm.

[0178] The quality evaluation score of the detail-enhanced image in the i-th iteration is the second NIQE score obtained by evaluating the quality of the detail-enhanced image in the i-th iteration using the NIQE algorithm.

[0179] Optionally, the satisfaction of the iteration condition can be determined as follows:

[0180] If the difference between the first NIQE score and the second NIQE score is less than or equal to zero or greater than or equal to a preset threshold, then the iteration condition is determined to be met.

[0181] If the difference between the first NIQE score and the second NIQE score is greater than zero and less than a preset threshold, then it is determined that the difference does not meet the iteration condition, and the detail enhancement image of the i-th iteration is directly output.

[0182] It should be noted that the NIQE score and the filter parameters are strongly correlated. Appropriate filter parameters can make the NIQE score have a local minimum on the original image or the enhanced image. Therefore, the filter parameters are updated based on the NIQE score to determine the most suitable filter parameters.

[0183] In practice, the total number of iterations is determined based on the difference between the average quality score of the original image and the image after enhancement. Each iteration changes the feature information of the original image. In other words, the feature information of the original image input is different each time. During the iterative enhancement process, by changing the feature information of the original image input each time, the quality of the image after enhancement using the feature information is made closer to the quality of the original image, ensuring that the image quality enhancement does not lead to a decrease in image quality.

[0184] like Figure 2 As shown, the iterative process of adaptive detail enhancement processing in this embodiment is as follows:

[0185] Step 200: Obtain the original image and determine the input image for detail enhancement processing related to the original image;

[0186] Optionally, the original image may include, but is not limited to, RGB images, HSV images, or images in other formats. This embodiment does not impose too many restrictions on the format of the original image.

[0187] It should be noted that the input image for detail enhancement processing in this embodiment includes, but is not limited to, any one or more of the following:

[0188] The original image; the V-channel image obtained by HSV conversion of the original image; the reconstructed image obtained by super-resolution processing of the original image; the contrast-enhanced image or overlay image obtained by contrast enhancement processing of the original image.

[0189] Step 201: For the input image of the detail enhancement processing, use at least one filtering function containing the filtering parameters updated in the i-th iteration to extract the feature information of the input image in the i-th iteration, and use the feature information of the i-th iteration to perform the detail enhancement processing on the input image in the i-th iteration to obtain the detail-enhanced image in the i-th iteration; wherein, the input image includes an image related to the original image;

[0190] Where i is an integer greater than or equal to 1;

[0191] In some embodiments, this embodiment provides a cascaded detail enhancement processing method. The specific processing steps involve using two filtering functions (a first filtering function and a second filtering function) for cascaded detail enhancement processing. The first filtering function contains first filtering parameters, and the second filtering function contains second filtering parameters. During the iterative detail enhancement processing, both the first and second filtering parameters are iteratively updated. In implementation, the original image is first enhanced using the first filtering function containing the first filtering parameters updated in the i-th iteration to obtain a first enhanced image. Then, the obtained first enhanced image is further enhanced using the second filtering function with the second filtering parameters updated in the i-th iteration. Through this cascaded enhancement processing method, and by utilizing the feature information corresponding to different filtering parameters to continuously enhance the details of the original image, the image quality enhancement effect can be improved, resulting in a clearer image. The specific implementation process is as follows:

[0192] It should be noted that the feature information in this embodiment corresponds one-to-one with the filtering parameters and filtering functions. The first filtering function extracts features from the original image to obtain first feature information, and the second filtering function extracts features from the first enhanced image to obtain second feature information. In some embodiments, the first filtering function is a guided filtering function, and the second filtering function is a Gaussian filtering function. The first feature information is used to represent the edge features of the original image, and the second feature information is used to represent the high-frequency information of the original image; or, the first filtering function is a Gaussian filtering function, and the second filtering function is a guided filtering function. The first feature information is used to represent the high-frequency features of the original image, and the second feature information is used to represent the edge information of the original image. The first filtering parameter represents the size of the first filtering window of the guided filtering function, and the second filtering parameter represents the size of the second filtering window of the Gaussian filtering function.

[0193] In implementation, at least one filtering function containing the filtering parameters updated in the i-th iteration is used to extract the feature information of the original image in the i-th iteration. The feature information from the i-th iteration is then used to perform detail enhancement processing on the original image in the i-th iteration to obtain the detail-enhanced image in the i-th iteration. Specifically, this includes:

[0194] Process 1-1: Using a first filtering function containing the first filtering parameters updated in the i-th iteration, extract the first feature information of the original image in the i-th iteration, and use the first feature information of the i-th iteration to perform detail enhancement processing on the original image in the i-th iteration to obtain the first enhanced image in the i-th iteration.

[0195] In implementation, the first feature information and the first filtering parameters in this embodiment are in one-to-one correspondence. The first feature information of the i-th iteration is superimposed on the original image to obtain the first enhanced image of the i-th iteration. Specifically, the first enhanced image is determined in the following way:

[0196] a) Filter the original image using a first filtering function that includes the first filtering parameters updated in the i-th iteration to obtain the first filtered image in the i-th iteration;

[0197] In some embodiments, the first filtering parameter can be the size of the first filtering window used for guided filtering. When guiding filtering is performed on the original image using the first filtering parameter, a guide map needs to be determined in advance. The guide map can be a separate image or the original image. The guide map is used to indicate where the smooth regions and edge regions are in the first filtering image. When the guide map is the input original image, guided filtering is used to implement a filtering operation that preserves the edges of the original image and can be applied to image reconstruction. In a specific implementation, the first feature map is determined in the following way:

[0198] a1) Based on the first filtering parameters and the guide map obtained by mean filtering of the original image, guide filtering is performed on the original image to obtain a guide filter map (i.e., the first filter map);

[0199] In practice, mean filtering is a typical linear filtering algorithm. From a frequency domain perspective, mean filtering is a low-pass filter that removes high-frequency signals, thus helping to eliminate sharp noise in images and achieve image smoothing and blurring. Ideally, mean filtering replaces each pixel in the image with the average value calculated from each pixel and its surrounding pixels (e.g., 4-neighborhood or 8-neighborhood). The sampling kernel data is typically a 3×3 matrix, calculating the value for each pixel in the image from left to right and top to bottom, ultimately yielding the processed image. Mean filtering can also incorporate two parameters: the number of iterations and the kernel size. Using the same kernel with multiple iterations will result in increasingly better performance. Similarly, with the same number of iterations, a larger kernel matrix leads to a more pronounced mean filtering effect. This embodiment does not impose excessive limitations on the specific kernel used for mean filtering or the number of iterations.

[0200] a2) Subtract the original image from the first filtered image of the i-th iteration to obtain the first feature information of the i-th iteration, that is, subtract the original image from the guided filtered image of the i-th iteration to obtain the first feature image containing the first feature information of the i-th iteration.

[0201] Since the guided filter image filters out the edge features of the original image, subtracting the guided filter image from the original image yields a first feature image containing the edge features of the original image. It should be noted that an image can be represented by a pixel matrix. Therefore, subtracting images specifically refers to subtracting the pixel matrices corresponding to different images, that is, subtracting the pixel matrix corresponding to the original image from the pixel matrix corresponding to the guided filter image.

[0202] This embodiment uses the filtered image obtained by mean filtering the original image as the guide image, so that the guided filtered image obtained by filtering the input original image can effectively extract edges, thereby enhancing the input original image and achieving the purpose of sharpening. The size of the filter window kernel of the guide filter can be updated. The larger the kernel, the stronger the enhancement effect on the original image. However, if the kernel becomes too large, it will easily lead to the distortion of the enhanced image.

[0203] b) The first feature information of the i-th iteration is superimposed on the original image to obtain the first enhanced image of the i-th iteration.

[0204] Since the first feature map contains the edge features of the original image, the edge features of the original image are enhanced by superimposing the first feature map onto the original image. It should be noted that an image can be represented by a pixel matrix; therefore, the superposition of images specifically refers to the superposition of the pixel matrices corresponding to different images, that is, the superposition of the pixel matrix corresponding to the original image and the pixel matrix corresponding to the first feature map.

[0205] Step 1-2: Using a second filtering function that includes the second filtering parameters updated in the i-th iteration, extract the second feature information of the first enhanced image in the i-th iteration, and use the second feature information of the i-th iteration to perform detail enhancement processing on the first enhanced image in the i-th iteration to obtain the detail enhanced image in the i-th iteration.

[0206] It should be noted that the second feature information corresponds to the second filtering parameters, which represent the size of the second filtering window for Gaussian filtering. The second enhancement process is actually performing Gaussian filtering on the first enhanced image and then using the high-frequency information extracted by the Gaussian filter to further enhance the first enhanced image. In practice, the image after this enhancement process is determined as follows:

[0207] c) Using a second filtering function containing the second filtering parameters updated in the i-th iteration, the first enhanced image of the i-th iteration is filtered to obtain the second filtered image of the i-th iteration.

[0208] In some embodiments, the second filtering parameter represents the size of the second filtering window of the Gaussian filter. It should be noted that the size of the filtering window in this embodiment specifically refers to the kernel size of the filtering window used.

[0209] In image processing, Gaussian filtering refers to combining frequency and time domain processing of an image, using it as a low-pass filter to obtain low-frequency information and then further filter out high-frequency information, thus smoothing the image. Gaussian filtering is a linear smoothing filter suitable for eliminating Gaussian noise and is widely used in image processing noise reduction. Gaussian filtering involves weighted averaging of the entire image; the value of each pixel is obtained by weighted averaging of its own value and the values ​​of its neighboring pixels. Specifically, Gaussian filtering scans each pixel in the image using a template (or convolution, mask), replacing the value of the center pixel of the template with the weighted average gray value of the pixels in the neighborhood defined by the template. Gaussian smoothing filters are very effective at suppressing noise that follows a normal distribution. In this embodiment, Gaussian filtering is used as a further enhancement detail extractor, effectively extracting some details of the image and enhancing image contrast. The size of the Gaussian filtering window is also updatable; a larger filter window results in a more significant enhancement effect, but when the filter window reaches a certain size, it can also lead to image distortion after enhancement. In practice, the second feature map is determined in the following way:

[0210] c1) Perform Gaussian filtering on the first enhancement map according to the second filtering parameters to obtain a Gaussian filtered map (i.e., the second filtered map);

[0211] It should be noted that Gaussian filtering is a low-pass filter that can obtain the low-frequency information of the first enhancement map, and then obtain the high-frequency information of the first enhancement map through the following steps, so as to further enhance the first enhancement map.

[0212] c2) Subtract the first enhanced map of the i-th iteration from the second filtered map of the i-th iteration to obtain the second feature map containing the second feature information of the i-th iteration.

[0213] Since the Gaussian filter image yields the low-frequency information of the first enhanced image, subtracting the Gaussian filter image from the first enhanced image yields the high-frequency information of the first enhanced image, i.e., the second feature information. Since an image can be represented by a pixel matrix, subtracting the Gaussian filter image from the first enhanced image actually yields a second feature image containing high-frequency information.

[0214] d) Superimpose the second feature information of the i-th iteration onto the first enhanced image of the i-th iteration to obtain the detail enhanced image of the i-th iteration, that is, superimpose the first enhanced image and the second feature image to obtain the detail enhanced image of the i-th iteration.

[0215] Since high-frequency information has already been extracted, superimposing the second feature map containing high-frequency information and the first enhancement map can further enhance the high-frequency information of the original image, making the enhanced image clearer.

[0216] In the above implementation process, optionally, the first filtering function is a Gaussian filtering function and the second filtering function is a guided filtering function.

[0217] It should be noted that the order of the first and second enhancement processes in the cascaded enhancement method of this embodiment is adjustable. As an optional implementation, it is shown below:

[0218] Process 2-1: Using a second filtering function containing the second filtering parameters updated in the i-th iteration, extract the second feature information of the original image in the i-th iteration, and use the second feature information of the i-th iteration to perform detail enhancement processing on the original image in the i-th iteration to obtain the first enhanced image in the i-th iteration;

[0219] Step 2-2: Using a first filtering function containing the first filtering parameters updated in the i-th iteration, extract the first feature information of the first enhanced image in the i-th iteration, and use the first feature information of the i-th iteration to perform detail enhancement processing on the first enhanced image in the i-th iteration to obtain the detail enhanced image in the i-th iteration.

[0220] The second filtering function is a Gaussian filtering function, and the first filtering function is a guided filtering function.

[0221] Either of the two processes described above can be chosen for implementation, and this embodiment does not impose excessive limitations on it. The specific implementation steps in processes 2-1 and 2-2 can be found in processes 1-1 and 1-2 above, and will not be repeated here.

[0222] Step 202: Determine whether the difference between the quality evaluation scores of the input image and the detail enhancement image in this iteration meets the iteration condition. If yes, proceed to step 203; otherwise, proceed to step 204.

[0223] Step 203: Update the filtering parameters of the i-th iteration to obtain the filtering parameters of the (i+1)-th iteration. Use at least one filtering function containing the filtering parameters of the (i+1)-th iteration to continue the detail enhancement processing of the input image, and return to step 202.

[0224] Step 204: Output the enhanced detail image for this iteration.

[0225] In some embodiments, if the original image is an RGB (Red (R), Green (G), Blue (B)) image, in order to avoid color changes after the original image is enhanced, this embodiment can also convert the original image into an HSV ((Hue, Saturation, Value), Hue (H), Saturation (S), Value (V)) image, and perform detail enhancement processing only on the V channel image in the HSV image.

[0226] In specific implementation, after acquiring the original image, before performing iterative detail enhancement processing on the original image using at least one filtering function containing iteratively updated filtering parameters, when the original image is determined to be an RGB image, the original image is converted to HSV to obtain an HSV original image. The luminance V channel image of the HSV original image is used as the input image for the detail enhancement processing. Iterative detail enhancement processing is performed on the luminance V channel image of the HSV original image using at least one filtering function containing iteratively updated filtering parameters. It should be noted that after replacing the original image with the V channel image, during the iterative enhancement process using the above method, the enhancement processing is not performed on the RGB image, but on the V channel image of the original image. Since no color channel enhancement processing is performed, the image quality enhancement method of this embodiment does not change the color of the original image, thereby ensuring that the color effect of the image obtained after image enhancement is not distorted.

[0227] In some embodiments, if the brightness V channel image of the original HSV image is subjected to iterative detail enhancement processing using at least one filtering function containing iteratively updated filtering parameters, then the detail-enhanced image of the last iteration is also a V channel image. Before displaying, the image of the last iteration of detail enhancement processing needs to be converted into an RGB detail-enhanced image, and then the RGB detail-enhanced image is displayed.

[0228] It should be noted that if the original image is replaced with the V channel image of an HSV image, the input V channel image is enhanced using the current feature information to obtain the enhanced image. The difference in quality evaluation score between the V channel image and the enhanced image is determined. If the iteration condition is met, the current filter parameters are updated, and the updated filter parameters are used to determine the updated feature information. The updated feature information is used as the feature information for the next iteration, and the enhancement process continues until the difference no longer meets the iteration condition. The image after the last enhancement is output. Since the original image is a V channel image, the image after the last enhancement is also a V channel image. Therefore, in order to restore the enhanced image to an RGB image for display, the image after the last enhancement is converted to an RGB image before outputting the RGB image.

[0229] like Figure 3 As shown, taking the original image as the input image for detail enhancement processing as an example, the implementation process of detail enhancement processing on the V channel image of the original image in this embodiment is as follows:

[0230] Step 300: Input the original image, convert the original image to an HSV original image, and obtain the V channel image of the HSV original image;

[0231] Step 301: Determine the feature information of this iteration using the filtering parameters of this iteration, and perform iterative detail enhancement processing on the V channel image using the feature information of this iteration to obtain the detail-enhanced image of this iteration;

[0232] In this context, there is a one-to-one correspondence between the filter parameters and the filter function, and also a one-to-one correspondence between the filter parameters and the feature information. This iteration represents the i-th iteration.

[0233] Step 302: Determine whether the difference between the quality evaluation scores of the V channel image and the detail enhancement image of this iteration meets the iteration condition. If yes, proceed to step 303; otherwise, proceed to step 304.

[0234] Step 303: Update the filter parameters for the current iteration to obtain the filter parameters for the next iteration, and return to execute step 301;

[0235] Step 304: Convert the detail enhancement image of this iteration into an RGB image and then display it.

[0236] In this embodiment, each enhancement process first uses various filtering parameters to determine various feature information, and then uses these feature information to enhance the original image, resulting in the enhanced image for each iteration. The difference between the enhancement processes lies in the different filtering parameters used, such as... Figure 4 As shown, the specific implementation process for a single detail enhancement process in this embodiment is as follows:

[0237] Step 400: Using the first filtering parameters of this iteration and the guide map obtained by mean filtering the V channel image of the original image, perform guide filtering on the original image to obtain the guide filter map;

[0238] Step 401: Subtract the original image from the guided filter image to obtain a first feature map containing the first feature information;

[0239] Step 402: Overlay the original image and the first feature map to obtain the first enhanced image;

[0240] Step 403: Perform Gaussian filtering on the first enhancement map according to the second filtering parameters of this iteration to obtain a Gaussian filtered map;

[0241] Step 404: Subtract the first enhanced image from the Gaussian filtered image to obtain a second feature image containing the second feature information;

[0242] Step 405: Overlay the first enhanced image and the second feature image to obtain the detail-enhanced image of this iteration.

[0243] In this case, the enhanced image is a V-channel image.

[0244] In some embodiments, before performing detail enhancement processing using the above method, the resolution of the original image can also be enhanced by using a super-resolution reconstruction algorithm (also known as super-resolution algorithm or super-resolution stretching) to reconstruct the resolution of the original image, thereby improving the sharpness of the original image. The specific implementation process of this method is as follows:

[0245] Using a super-resolution algorithm, the resolution of the input image for super-resolution processing is reconstructed to a preset resolution to obtain a reconstructed image, wherein the preset resolution is higher than the resolution of the original image; the reconstructed image is used as the input image for the detail enhancement processing.

[0246] The input image for the super-resolution processing is either the original image or the contrast-enhanced image, wherein the contrast-enhanced image is obtained by performing contrast enhancement processing on the original image.

[0247] Super-resolution algorithms are techniques used to restore details in images when they are enlarged. They utilize knowledge from fields such as digital image processing and computer vision, employing specific algorithms and processing flows to reconstruct a high-resolution image from a given low-resolution image. Super-resolution algorithms include two main categories: traditional algorithms and deep learning algorithms.

[0248] Currently, traditional super-resolution algorithms mainly rely on basic digital image processing techniques for reconstruction. For example, interpolation-based super-resolution reconstruction involves fitting known pixel information to unknown pixel information on a plane, typically accomplished by a predefined transformation function or interpolation kernel. While interpolation-based methods are computationally simple and easy to understand, they also have some significant drawbacks, often leading to blurring and jagged edges in the reconstructed image. Common interpolation-based methods include nearest-neighbor interpolation, bilinear interpolation, and bicubic interpolation.

[0249] This embodiment employs deep learning algorithms for super-resolution reconstruction. Through multi-layer nonlinear transformations, it extracts high-level abstract features of the image and learns the potential distribution patterns of the image, thereby acquiring the ability to make reasonable judgments or predictions about the reconstructed image. The super-resolution algorithms used in this embodiment include, but are not limited to, at least one of SRCNN (Super Resolution Convolutional Network), VDSR (Very Deep Convolutional Networks Super Resolution), EDSR (Enhanced Deep Super-Resolution Network), and SRGAN (Super Resolution Generative Adversarial Network).

[0250] In some embodiments, the preset resolution in this embodiment includes, but is not limited to, at least one of 4K resolution and 8K resolution, wherein the preset resolution is much higher than the resolution of the original image. The specific super-resolution reconstruction process is as follows:

[0251] First, a super-resolution factor corresponding to the size of the original image is determined; second, the resolution of the original image is reconstructed to a preset resolution based on the super-resolution factor and the super-resolution algorithm.

[0252] In this process, multiple super-resolution scaling factors can be preset. The smaller the size, the larger the super-resolution scaling factor corresponding to that size. Super-resolution reconstruction is performed using the super-resolution scaling factor corresponding to the size of the original image. Thus, the super-resolution scaling factor is dynamically determined based on the input original image, which determines the stretching ratio of the input original image, thereby balancing the effect and response time of the super-resolution algorithm.

[0253] like Figure 5 As shown, taking a preset resolution of 4K as an example, the super-resolution reconstruction method in this embodiment is explained as follows:

[0254] Step 500: Obtain the original image and determine the maximum value of the width and height of the original image;

[0255] Step 501: Divide 3840 by the maximum value to obtain parameter D;

[0256] Optionally, the 3840 mentioned above can be replaced with a pixel value within the 4K resolution range, and this embodiment does not impose too many limitations on this;

[0257] Step 502: Calculate the absolute value of the difference between parameter D and 1, 2, 3, 4, and 8 respectively to obtain the corresponding D1, D2, D3, D4, and D8;

[0258] The selection of 1, 2, 3, 4, and 8 can be determined according to the requirements. This is only an example to determine how close D is to 1K, 2K, 3K, 4K, and 8K resolutions, respectively. The values ​​selected based on this judgment principle are all within the protection scope of this disclosure, and will not be listed one by one here.

[0259] Step 503: Select the minimum value from D1, D2, D3, D4, and D8, and determine the super-resolution factor corresponding to the minimum value;

[0260] In practice, if D1 is the smallest, it means that the resolution of the original image is close to 4K. The super-resolution factor corresponding to D1 is 1, that is, the original image is stretched to 4K by directly using the bicubic interpolation method.

[0261] If D2 is the smallest, it means that the resolution of the original image is close to 2K. The super-resolution factor corresponding to D2 is 2. Using a super-resolution algorithm of 2, the original image is stretched to 4K.

[0262] If D3 is the smallest, it means that the resolution of the original image is close to 1.5K. The super-resolution factor corresponding to D3 is 3. Using a super-resolution algorithm of 3 times, the original image is stretched to 4K.

[0263] If D4 is the smallest, it means that the resolution of the original image is close to 1K. The super-resolution factor corresponding to D4 is 4. Using a super-resolution algorithm of 4x, the original image is stretched to 4K.

[0264] If D5 is the smallest, it means that the resolution of the original image is close to 0.5K. The super-resolution factor corresponding to D5 is 8. Using the 8x super-resolution algorithm, the original image is stretched to 4K.

[0265] Step 504: Using the super-resolution factor and super-resolution algorithm, reconstruct the resolution of the original image to 4K resolution.

[0266] After reconstructing the original image to a preset resolution using the above method, iterative detail enhancement processing is performed on the reconstructed image. This involves using the reconstructed image as input for detail enhancement. Thus, after reconstructing the resolution of the original image, further detail enhancement processing is applied to the reconstructed image, effectively improving the clarity of the original image. This is particularly suitable for the display needs of ultra-large-sized display devices, providing users with an ultra-clear video viewing experience.

[0267] In some embodiments, to further enhance the image quality of the original image, the contrast of the original image can also be enhanced. It should be noted that in this embodiment, there is no prescribed order between the super-resolution reconstruction of the original image and the contrast enhancement. Specifically, a super-resolution algorithm can be used first to reconstruct the original image to a preset resolution, obtaining a reconstructed image. Then, contrast enhancement processing is performed on the reconstructed image. Finally, various feature information is used to further enhance the image obtained after contrast enhancement. In some embodiments, contrast enhancement processing can also be performed on the original image first, super-resolution processing can be applied to the obtained contrast-enhanced image, and finally, detail enhancement processing can be performed on the reconstructed image obtained after super-resolution processing.

[0268] In some embodiments, detail enhancement processing is performed on the original image. Specifically, detail enhancement processing is performed on the V channel image of the original HSV image to obtain a detail-enhanced image, which is then converted into an RGB detail-enhanced image. Similarly, contrast enhancement processing is performed on the original image. Specifically, contrast enhancement processing is performed on the V channel image of the original HSV image to obtain a contrast-enhanced image or a superimposed image, which is then converted into an RGB image.

[0269] In some embodiments, the principle of contrast enhancement processing in this embodiment is to continuously perform contrast enhancement processing on the reconstructed image after super-resolution processing or the original image without super-resolution processing through iterative gamma transformation. The gamma value is updated during the iteration process. In each iteration, the gamma value is used to perform gamma transformation on the reconstructed image or the original image, and the contrast difference between the transformed image and the original image is calculated. If the contrast difference is greater than the contrast threshold, the gamma value needs to be updated and the updated gamma value is used to continue to perform gamma transformation on the reconstructed image or the original image. When the contrast difference between the gamma-transformed image and the original image is less than or equal to the contrast threshold, it is determined that the gamma-transformed image is an enhanced image and can be output to the display device for display.

[0270] The implementation process of the contrast enhancement processing scheme in this embodiment is described below:

[0271] The reconstructed image after super-resolution processing or the original image without super-resolution processing is used as the input image, and iterative contrast enhancement processing is performed using the iteratively updated gamma value:

[0272] The input image for the contrast enhancement process is subjected to gamma transformation according to the gamma value updated in the j-th iteration to obtain the contrast-enhanced image in the j-th iteration; where j is an integer greater than or equal to 1.

[0273] If the contrast difference between the contrast-enhanced image of the j-th iteration and the contrast-enhanced input image is greater than the contrast threshold, the contrast of the contrast-enhanced input image is replaced with the contrast of the contrast-enhanced image of the j-th iteration, and the gamma value of the j-th iteration is updated to obtain the gamma value of the (j+1)-th iteration. Using the gamma value of the (j+1)-th iteration, the contrast-enhanced input image is further subjected to iterative contrast enhancement processing until the contrast difference between the contrast-enhanced image of the j-th iteration and the contrast-enhanced input image is less than or equal to the contrast threshold.

[0274] The contrast-enhanced image from the last iteration is used as the input image for the detail enhancement process.

[0275] In practice, the gamma value of the j-th iteration is used to perform a gamma transformation on the original image to obtain the contrast-enhanced image of the j-th iteration. It should be noted that the gamma value of the first iteration is preset, and the gamma value used when performing gamma transformation on the original image is different for different iterations. That is, by continuously updating the gamma value, gamma transformation with different gamma values ​​is continuously performed on the original image, so that the contrast of the transformed image can be close to the contrast of the original image.

[0276] Contrast ratio is used to represent the degree of stretching between bright and dark areas, and usually reflects the clarity of the image quality. In this embodiment, the contrast ratio of the original image and the gamma-transformed image can be calculated using the following formula (1):

[0277] C=∑ δ δ(i,j) 2 P δ (i,j) Formula (1);

[0278] Where δ(i,j)=|ij| represents the gray value difference between pixel i and its neighboring pixel j, P δ (i,j) represents the probability that the pixel difference is δ.

[0279] The neighboring pixels can be selected as 4 neighboring pixels (top, bottom, left, right) or 8 neighboring pixels (top, bottom, left, right, top left, top right, bottom left, bottom right).

[0280] If the contrast difference between the contrast-enhanced image of the j-th iteration and the contrast-enhanced input image is greater than a contrast threshold, the contrast of the input image is replaced with the contrast of the contrast-enhanced image of the j-th iteration, and the gamma value of the j-th iteration is updated to obtain the gamma value of the (j+1)-th iteration. Gamma transformation is then performed on the original image until the contrast difference is determined to be less than or equal to the contrast threshold. The contrast-enhanced image of the last iteration is used as the input image for the detail enhancement processing.

[0281] In practice, the iterative process involves updating the gamma value and performing a gamma transform using the updated gamma value. The contrast of the input image is also continuously updated along with the gamma value. After each transform, the contrast of the input image is replaced with the contrast of the transformed image. The difference between the contrast of the image obtained after the previous transform and the contrast obtained after the current transform is continuously calculated. If the difference is not greater than a threshold, indicating that the contrast improvement is insufficient, the iteration can be stopped. The contrast-enhanced image from the last iteration is output.

[0282] It should be noted that there are various methods for contrast enhancement, including logarithmic transformation, power-law transformation, and histogram equalization. In this embodiment, considering the controllability and adjustability of power-law transformation, power-law transformation is chosen for contrast enhancement. Power-law transformation, also known as gamma transformation, is a power function operation performed on each pixel, as shown in formula (2).

[0283] s = cr γ Formula (2);

[0284] Where c and γ (i.e. gamma) are constants, r represents the pixel value, and s represents the pixel value after gamma transformation.

[0285] When γ < 1, it can increase image brightness and decrease image contrast. When γ > 1, it can decrease image brightness, making the image darker, but increasing contrast. Therefore, to appropriately increase image contrast, a γ value greater than 1 should be used.

[0286] It should be noted that, in order to ensure that the colors of the original image are not changed during the gamma transformation, the input image can be transformed into an HSV image, and the V channel image of the HSV image is subjected to gamma transformation. Specifically, when it is determined that the reconstructed image after super-resolution processing is an RGB image, the reconstructed image is converted to HSV to obtain the original HSV image, and the V channel image of the original HSV image is used as the input image for the contrast enhancement processing; or, when it is determined that the original image without super-resolution processing is an RGB image, the original image is converted to HSV to obtain the original HSV image, and the V channel image of the original HSV image is used as the input image for the contrast enhancement processing.

[0287] In practice, a gamma transform can be performed on the V channel image to calculate the contrast of the V channel image. It can be determined whether the increase in contrast is greater than the contrast threshold. If so, the gamma (γ) value is increased, a gamma transform is performed, the contrast is iteratively calculated, the increase in contrast is determined again, and finally the γ value is determined.

[0288] It should be noted that if a gamma transform is performed on the V channel of the original image, the final gamma-transformed image will also be a V channel image. Therefore, it is necessary to convert the V channel image to an RGB image.

[0289] In some embodiments, since the contrast of the original image is improved after gamma transformation, the overall brightness of the gamma-transformed image will decrease. In this embodiment, after determining that the contrast difference is less than or equal to the contrast threshold, brightness stretching is first performed, and the brightness-stretched image is used as the final output image. In practice, brightness stretching is performed in the following manner:

[0290] Determine the difference between the average brightness of the original image and the contrast-enhanced image obtained by the last gamma transform; wherein, if the input image is a V-channel image, the average brightness of the input image is the average of all gray values ​​in the V-channel image. Similarly, if the input image is a V-channel image, the contrast-enhanced image obtained by the last gamma transform is also a V-channel image, and the average brightness of the contrast-enhanced image obtained by the last gamma transform is also the average of all gray values.

[0291] The difference is superimposed on the brightness of the contrast-enhanced image obtained by the last gamma transform to obtain a superimposed image; wherein, the brightness of the image is taken as a brightness matrix, and the difference is superimposed on the brightness of the contrast-enhanced image obtained by the gamma transform, that is, the difference is superimposed on the brightness matrix to obtain a superimposed image.

[0292] The contrast-enhanced image obtained from the last gamma transform is replaced with the superimposed image, which can be considered as an image after brightness stretching. By superimposing the difference in the mean brightness and stretching the brightness range, the brightness of the image can be stabilized. It should be noted that after brightness stretching, since the value range of each pixel may not be within the standard range (e.g., 0-255), the value of each pixel can be mapped to the standard range to finally obtain the superimposed image.

[0293] In practice, by superimposing brightness, the overall brightness of the image can remain unchanged, and by enhancing contrast, the image clarity can be effectively improved. Since this solution maintains brightness on the basis of contrast enhancement, it can achieve a very good enhancement effect.

[0294] In some embodiments, if the input original image is a V-channel image, then the image after brightness stretching, i.e., the overlay image, is also a V-channel image. Therefore, the overlay image needs to be converted to an RGB image before output.

[0295] like Figure 6 As shown in the figure, this embodiment provides a contrast enhancement processing method, and the specific implementation process of the method is as follows:

[0296] Step 600: Convert the reconstructed image after super-resolution processing or the original image without super-resolution processing to HSV to obtain an HSV image, and obtain the V channel image of the HSV image;

[0297] Step 601: Using the gamma value of this iteration, perform gamma transformation on the V channel image to obtain the contrast-enhanced image after this transformation;

[0298] Here, "this time" refers to the j-th iteration.

[0299] Step 602: Determine whether the contrast difference between the transformed image and the V channel image is greater than the contrast threshold. If yes, proceed to step 603; otherwise, proceed to step 604.

[0300] Step 603: Replace the contrast of the V channel image with the contrast of the contrast-enhanced image after this transformation, and update the gamma value of this iteration to obtain the gamma value of the next iteration, and return to execute step 601.

[0301] In practice, the gamma value can be updated by either increasing the gamma value by adding a set step size or decreasing the gamma value by subtracting a set step size. The method can be determined according to actual needs, and this embodiment does not impose too many restrictions on this.

[0302] Step 604: Determine the difference in the average brightness of the V channel image and the contrast-enhanced image obtained by the last gamma transform;

[0303] Step 605: Superimpose the difference with the brightness of the contrast-enhanced image obtained by the last gamma transform to obtain a superimposed image;

[0304] Step 606: Convert the superimposed image into an RGB image and output it.

[0305] To provide a complete explanation of the three enhancement methods provided in this embodiment, this embodiment offers a process flow for enhancing image quality. This flow is merely an example, used to illustrate the specific process involving the three enhancement methods, such as... Figure 7 As shown, the specific implementation method of this process is as follows:

[0306] Step 700: Input the original image, and use a super-resolution algorithm to reconstruct the resolution of the original image to the preset resolution to obtain the reconstructed image;

[0307] Step 701: Convert the reconstructed image into an HSV image and obtain the V channel image of the HSV image;

[0308] Step 702: Perform gamma transformation on the V channel image according to the gamma value updated in the j-th iteration to obtain the contrast-enhanced image of the j-th iteration; determine the contrast difference between the contrast-enhanced image of the j-th iteration and the V channel image. If the contrast difference is greater than the contrast threshold, replace the contrast of the V channel image with the contrast of the contrast-enhanced image of the j-th iteration, and update the gamma value of the j-th iteration to obtain the gamma value of the (j+1)-th iteration. Use the gamma value of the (j+1)-th iteration to continue iterative contrast enhancement processing on the V channel image until the contrast difference between the contrast-enhanced image of the j-th iteration and the V channel image is less than or equal to the contrast threshold.

[0309] Where j is an integer greater than or equal to 1;

[0310] Step 703: Determine the difference between the average brightness of the V channel image and the contrast-enhanced image of the last iteration, and superimpose the difference with the brightness of the contrast-enhanced image of the last iteration to obtain a superimposed image;

[0311] Step 704: Using at least one filtering function containing the filtering parameters updated in the i-th iteration, extract the feature information of the overlay image in the i-th iteration; use the feature information of the i-th iteration to perform detail enhancement processing on the overlay image in the i-th iteration to obtain the detail-enhanced image in the i-th iteration; when it is determined that the difference in quality evaluation scores between the overlay image and the detail-enhanced image in the i-th iteration satisfies the iteration condition, update the filtering parameters of the i-th iteration to obtain the filtering parameters of the (i+1)-th iteration; use at least one filtering function containing the filtering parameters of the (i+1)-th iteration to continue performing detail enhancement processing on the overlay image until the difference in quality evaluation scores between the overlay image and the detail-enhanced image in the (i+1)-th iteration satisfies the iteration condition, then stop.

[0312] Where i is an integer greater than or equal to 1;

[0313] Step 705: Display the enhanced detail image from the last iteration.

[0314] In summary, this disclosure provides three enhancement processing methods: First, detail enhancement processing, which utilizes the edge feature information and high-frequency information of the original image to enhance the details of the input image; second, super-resolution reconstruction processing, which uses a super-resolution reconstruction algorithm to perform super-resolution reconstruction on the input image; and third, contrast enhancement processing, which uses gamma transform to enhance the contrast of the input image. In specific implementations, these three enhancement processing methods can be implemented independently or in combination; this embodiment does not impose excessive limitations on this.

[0315] In this embodiment, the order of the three enhancement processing methods described above can be defined as needed, and this embodiment does not impose excessive limitations on this. In some embodiments, the original image is first subjected to super-resolution processing to obtain a reconstructed image, then the V channel image of the reconstructed HSV image is subjected to iterative contrast enhancement processing to obtain an RGB superimposed image, and finally, the V channel image of the superimposed HSV image is subjected to iterative detail enhancement processing to obtain an RGB detail-enhanced image. In some embodiments, the V channel image of the original HSV image is first subjected to iterative contrast enhancement processing to obtain an RGB superimposed image, then the RGB superimposed image is subjected to super-resolution reconstruction processing (super-resolution processing) to obtain a reconstructed image, and finally, the V channel image of the reconstructed HSV image is subjected to iterative detail enhancement processing to obtain an RGB detail-enhanced image.

[0316] Based on the same inventive concept, this disclosure also provides an adaptive image quality enhancement method. Since this method is the same as the method used by the display device in this disclosure, and the principle of solving the problem is similar to that of the display device, the implementation of this method can be referred to the implementation of the display device, and repeated details will not be repeated.

[0317] like Figure 8 As shown in the figure, this embodiment also provides a method for adaptive image quality enhancement, and the specific implementation process of the method is as follows:

[0318] Step 800: Obtain the original image and perform iterative detail enhancement processing on the original image using at least one filtering function containing iteratively updated filtering parameters; specifically, this includes: extracting the feature information of the input image for the detail enhancement processing in the i-th iteration using at least one filtering function containing filtering parameters updated in the i-th iteration, and performing the i-th iteration detail enhancement processing on the input image using the feature information of the i-th iteration to obtain the detail-enhanced image in the i-th iteration; wherein, the input image includes an image related to the original image, and i is an integer greater than or equal to 1;

[0319] In practice, the input original image is enhanced using the various feature information obtained in this process to obtain the enhanced image. The various feature information is determined based on the various filtering parameters used in this process. The filtering parameters are used to represent the size of the filtering window used when extracting feature information from the original image.

[0320] Step 801: When the difference between the quality evaluation scores of the input image and the detail-enhanced image of the i-th iteration satisfies the iteration condition, the filtering parameters of the i-th iteration are updated to obtain the filtering parameters of the (i+1)-th iteration. The input image is then further enhanced with at least one filtering function containing the filtering parameters of the (i+1)-th iteration until the difference between the quality evaluation scores of the input image and the detail-enhanced image of the (i+1)-th iteration satisfies the iteration condition.

[0321] Step 802: Display the enhanced detail image from the last iteration.

[0322] As one possible implementation, after acquiring the original image and before performing iterative detail enhancement processing on the original image using at least one filtering function containing iteratively updated filtering parameters, the method further includes:

[0323] When the original image is determined to be an RGB image, the original RGB image is converted to HSV color channels to obtain the original HSV image;

[0324] The brightness V channel image of the original HSV image is used as the input image for the detail enhancement process.

[0325] As one possible implementation, before displaying the image from the last iteration of detail enhancement processing, the method further includes:

[0326] Convert the image processed in the last iteration of detail enhancement to an RGB detail-enhanced image;

[0327] The detailed enhanced image showing the last iteration specifically includes:

[0328] Display the RGB detail-enhanced image.

[0329] As one possible implementation, the step of extracting feature information of the original image in the i-th iteration using at least one filtering function containing the filtering parameters updated in the i-th iteration, and then using the feature information of the i-th iteration to perform detail enhancement processing on the original image in the i-th iteration to obtain a detail-enhanced image in the i-th iteration specifically includes:

[0330] Using a first filtering function that includes the first filtering parameters updated in the i-th iteration, the first feature information of the original image in the i-th iteration is extracted. The first feature information of the i-th iteration is then used to perform detail enhancement processing on the original image in the i-th iteration to obtain the first enhanced image in the i-th iteration.

[0331] Using a second filtering function that includes the second filtering parameters updated in the i-th iteration, the second feature information of the first enhanced image in the i-th iteration is extracted. The first enhanced image is then subjected to detail enhancement processing in the i-th iteration using the second feature information of the i-th iteration, resulting in the detail-enhanced image in the i-th iteration.

[0332] As one possible implementation, the original image is subjected to detail enhancement processing using the first feature information from the i-th iteration to obtain the first enhanced image of the i-th iteration, including:

[0333] The first feature information of the i-th iteration is superimposed on the original image to obtain the first enhanced image of the i-th iteration;

[0334] The step of performing detail enhancement processing on the first enhanced image using the second feature information from the i-th iteration includes:

[0335] The second feature information of the i-th iteration is superimposed on the first enhanced image of the i-th iteration to obtain the detail enhanced image of the i-th iteration.

[0336] As one possible implementation, extracting the first feature information of the original image in the i-th iteration using a first filtering function that includes the first filtering parameters updated in the i-th iteration includes:

[0337] The original image is filtered using a first filtering function that includes the first filtering parameters updated in the i-th iteration to obtain the first filtered image in the i-th iteration; the original image is subtracted from the first filtered image in the i-th iteration to obtain the first feature information in the i-th iteration.

[0338] The step of extracting the second feature information of the first enhanced image in the i-th iteration using a second filtering function that includes the second filtering parameters updated in the i-th iteration includes:

[0339] The first enhanced image of the i-th iteration is filtered using a second filtering function that includes the second filtering parameters updated in the i-th iteration to obtain the second filtered image of the i-th iteration; the first enhanced image of the i-th iteration is subtracted from the second filtered image of the i-th iteration to obtain the second feature information of the i-th iteration.

[0340] As one possible implementation, the first filtering function is a guided filtering function, and the second filtering function is a Gaussian filtering function; or,

[0341] The first filtering function is a Gaussian filtering function, and the second filtering function is a guided filtering function.

[0342] As one possible implementation, the quality assessment score of the original image is a first NIQE score obtained by evaluating the quality of the original image using the No Reference Image Quality Assessment (NIQE) algorithm.

[0343] The quality evaluation score of the detail-enhanced image in the i-th iteration is the second NIQE score obtained by evaluating the quality of the detail-enhanced image in the i-th iteration using the NIQE algorithm.

[0344] As one possible implementation, determining that the difference between the quality evaluation scores of the original image and the detail-enhanced image of the i-th iteration satisfies the iteration condition includes:

[0345] The difference between the first NIQE score and the second NIQE score is determined to be less than or equal to zero or greater than or equal to a preset threshold.

[0346] As one possible implementation, after acquiring the original image and before performing iterative detail enhancement processing on the original image using at least one filtering function containing iteratively updated filtering parameters, the method further includes:

[0347] Using a super-resolution algorithm, the resolution of the input image for super-resolution processing is reconstructed to a preset resolution to obtain a reconstructed image, wherein the preset resolution is higher than the resolution of the original image; wherein the input image for super-resolution processing is either the original image or a contrast-enhanced image, and the contrast-enhanced image is obtained by performing contrast enhancement processing on the original image;

[0348] The reconstructed image is used as the input image for the detail enhancement process.

[0349] As one possible implementation, the step of using a super-resolution algorithm to reconstruct the resolution of the original image to a preset resolution includes:

[0350] The super-resolution factor corresponding to the size is determined based on the size of the original image;

[0351] Based on the super-resolution factor and the super-resolution algorithm, the resolution of the original image is reconstructed to a preset resolution.

[0352] As one possible implementation, after acquiring the original image and before performing iterative detail enhancement processing on the original image using at least one filtering function containing iteratively updated filtering parameters, the method further includes:

[0353] The reconstructed image after super-resolution processing or the original image without super-resolution processing is used as the input image, and iterative contrast enhancement processing is performed using the iteratively updated gamma value:

[0354] The input image for the contrast enhancement process is subjected to gamma transformation according to the gamma value updated in the j-th iteration to obtain the contrast-enhanced image in the j-th iteration; where j is an integer greater than or equal to 1.

[0355] If the contrast difference between the contrast-enhanced image of the j-th iteration and the contrast-enhanced input image is greater than the contrast threshold, the contrast of the contrast-enhanced input image is replaced with the contrast of the contrast-enhanced image of the j-th iteration, and the gamma value of the j-th iteration is updated to obtain the gamma value of the (j+1)-th iteration. Using the gamma value of the (j+1)-th iteration, the contrast-enhanced input image is further subjected to iterative contrast enhancement processing until the contrast difference between the contrast-enhanced image of the j-th iteration and the contrast-enhanced input image is less than or equal to the contrast threshold.

[0356] The contrast-enhanced image from the last iteration is used as the input image for the detail enhancement process.

[0357] As one possible implementation, before using the contrast-enhanced image from the last iteration as the input image for the detail enhancement process, the method further includes:

[0358] Determine the difference in the mean brightness between the input image of the contrast enhancement process and the contrast enhancement image of the last iteration;

[0359] The difference is superimposed with the brightness of the contrast-enhanced image from the last iteration to obtain a superimposed image;

[0360] The overlaid image is used as the input image for the detail enhancement process.

[0361] As one possible implementation, prior to the iterative contrast enhancement processing using iteratively updated gamma values, the method further includes:

[0362] When the reconstructed image after super-resolution processing is determined to be an RGB image, the reconstructed image is converted to HSV to obtain the original HSV image, and the V channel image of the original HSV image is used as the input image for the contrast enhancement processing; or,

[0363] When the original image that has not undergone super-resolution processing is determined to be an RGB image, the original image is converted to HSV to obtain an original HSV image, and the V channel image of the original HSV image is used as the input image for the contrast enhancement processing.

[0364] As one possible implementation, before using the overlaid image as the input image for the detail enhancement process, the method further includes:

[0365] The overlay image is converted into an RGB overlay image.

[0366] Based on the same inventive concept, this disclosure also provides an adaptive image quality enhancement device. Since this device is the same as the device corresponding to the method in this disclosure, and the principle of the device in solving the problem is similar to that of the method, the implementation of the device can refer to the implementation of the method, and the repeated parts will not be described again.

[0367] like Figure 9 As shown, the device includes:

[0368] The enhancement processing unit 900 is used to acquire an original image and perform iterative detail enhancement processing on the original image using at least one filtering function containing iteratively updated filtering parameters. Specifically, it includes: extracting feature information of the original image in the i-th iteration from the input image of the detail enhancement processing using at least one filtering function containing filtering parameters updated in the i-th iteration; and performing detail enhancement processing on the original image in the i-th iteration using the feature information of the i-th iteration to obtain a detail-enhanced image in the i-th iteration; where i is an integer greater than or equal to 1.

[0369] The difference judgment unit 901 is used to determine that when the difference between the quality evaluation scores of the original image and the detail enhancement image of the i-th iteration meets the iteration condition, update the filtering parameters of the i-th iteration to obtain the filtering parameters of the (i+1)-th iteration, and continue to perform detail enhancement processing on the detail enhancement image of the i-th iteration using at least one filtering function containing the filtering parameters of the (i+1)-th iteration until the difference between the quality evaluation scores of the original image and the detail enhancement image of the (i+1)-th iteration meets the iteration condition.

[0370] Enhanced output unit 902 is used to display a detailed enhanced image of the last iteration.

[0371] As one possible implementation, before displaying the image from the last iteration of detail enhancement processing, a transformation unit is further included specifically for:

[0372] Convert the image processed in the last iteration of detail enhancement to an RGB detail-enhanced image;

[0373] The enhanced output unit 902 is further used for:

[0374] Display the RGB detail-enhanced image.

[0375] As one possible implementation, the enhancement processing unit 900 is specifically used for:

[0376] Using a first filtering function that includes the first filtering parameters updated in the i-th iteration, the first feature information of the original image in the i-th iteration is extracted. The first feature information of the i-th iteration is then used to perform detail enhancement processing on the original image in the i-th iteration to obtain the first enhanced image in the i-th iteration.

[0377] Using a second filtering function that includes the second filtering parameters updated in the i-th iteration, the second feature information of the first enhanced image in the i-th iteration is extracted. The first enhanced image is then subjected to detail enhancement processing in the i-th iteration using the second feature information of the i-th iteration, resulting in the detail-enhanced image in the i-th iteration.

[0378] As one possible implementation, the enhancement processing unit 900 is specifically used for:

[0379] The first feature information of the i-th iteration is superimposed on the original image to obtain the first enhanced image of the i-th iteration;

[0380] The second feature information of the i-th iteration is superimposed on the first enhanced image of the i-th iteration to obtain the detail enhanced image of the i-th iteration.

[0381] As one possible implementation, the enhancement processing unit 900 is specifically used for:

[0382] The original image is filtered using a first filtering function that includes the first filtering parameters updated in the i-th iteration to obtain the first filtered image in the i-th iteration; the original image is subtracted from the first filtered image in the i-th iteration to obtain the first feature information in the i-th iteration.

[0383] The first enhanced image of the i-th iteration is filtered using a second filtering function that includes the second filtering parameters updated in the i-th iteration to obtain the second filtered image of the i-th iteration; the first enhanced image of the i-th iteration is subtracted from the second filtered image of the i-th iteration to obtain the second feature information of the i-th iteration.

[0384] As one possible implementation, the first filtering function is a guided filtering function, and the second filtering function is a Gaussian filtering function; or,

[0385] The first filtering function is a Gaussian filtering function, and the second filtering function is a guided filtering function.

[0386] As one possible implementation, the quality assessment score of the original image is a first NIQE score obtained by evaluating the quality of the original image using the No Reference Image Quality Assessment (NIQE) algorithm.

[0387] The quality evaluation score of the detail-enhanced image in the i-th iteration is the second NIQE score obtained by evaluating the quality of the detail-enhanced image in the i-th iteration using the NIQE algorithm.

[0388] As one possible implementation, the difference determination unit 901 is specifically used for:

[0389] The difference between the first NIQE score and the second NIQE score is determined to be less than or equal to zero or greater than or equal to a preset threshold.

[0390] As one possible implementation, after acquiring the original image and before performing iterative detail enhancement processing on the original image using at least one filtering function containing iteratively updated filtering parameters, a reconstruction unit is further included, specifically for:

[0391] Using a super-resolution algorithm, the resolution of the input image for super-resolution processing is reconstructed to a preset resolution to obtain a reconstructed image, wherein the preset resolution is higher than the resolution of the original image; wherein the input image for super-resolution processing is either the original image or a contrast-enhanced image, and the contrast-enhanced image is obtained by performing contrast enhancement processing on the original image;

[0392] The reconstructed image is used as the input image for the detail enhancement process.

[0393] As one possible implementation, the reconstruction unit is specifically used for:

[0394] The super-resolution factor corresponding to the size is determined based on the size of the original image;

[0395] Based on the super-resolution factor and the super-resolution algorithm, the resolution of the original image is reconstructed to a preset resolution.

[0396] As one possible implementation, after acquiring the original image and before performing iterative detail enhancement processing on the original image using at least one filtering function containing iteratively updated filtering parameters, a contrast enhancement unit is further included, specifically for:

[0397] The reconstructed image after super-resolution processing or the original image without super-resolution processing is used as the input image, and iterative contrast enhancement processing is performed using the iteratively updated gamma value:

[0398] The input image for the contrast enhancement process is subjected to gamma transformation according to the gamma value updated in the j-th iteration to obtain the contrast-enhanced image in the j-th iteration; where j is an integer greater than or equal to 1.

[0399] If the contrast difference between the contrast-enhanced image of the j-th iteration and the contrast-enhanced input image is greater than the contrast threshold, the contrast of the contrast-enhanced input image is replaced with the contrast of the contrast-enhanced image of the j-th iteration, and the gamma value of the j-th iteration is updated to obtain the gamma value of the (j+1)-th iteration. Using the gamma value of the (j+1)-th iteration, the contrast-enhanced input image is further subjected to iterative contrast enhancement processing until the contrast difference between the contrast-enhanced image of the j-th iteration and the contrast-enhanced input image is less than or equal to the contrast threshold.

[0400] The contrast-enhanced image from the last iteration is used as the input image for the detail enhancement process.

[0401] As one possible implementation, before using the contrast-enhanced image from the last iteration as the input image for the detail enhancement process, the contrast enhancement unit is further configured to:

[0402] Determine the difference in the mean brightness between the input image of the contrast enhancement process and the contrast enhancement image of the last iteration;

[0403] The difference is superimposed with the brightness of the contrast-enhanced image from the last iteration to obtain a superimposed image;

[0404] The overlaid image is used as the input image for the detail enhancement process.

[0405] As one possible implementation, before the iterative contrast enhancement processing using the iteratively updated gamma value, the contrast enhancement unit is further configured to:

[0406] When the reconstructed image after super-resolution processing is determined to be an RGB image, the reconstructed image is converted to HSV to obtain the original HSV image, and the V channel image of the original HSV image is used as the input image for the contrast enhancement processing; or,

[0407] When the original image that has not undergone super-resolution processing is determined to be an RGB image, the original image is converted to HSV to obtain an original HSV image, and the V channel image of the original HSV image is used as the input image for the contrast enhancement processing.

[0408] As one possible implementation, before using the overlaid image as the input image for the detail enhancement process, the contrast enhancement unit is further configured to:

[0409] The overlay image is converted into an RGB overlay image.

[0410] Based on the same inventive concept, this disclosure also provides a non-transient computer storage medium storing a computer program thereon, wherein the program, when executed by a processor, performs the following steps:

[0411] The process involves acquiring an original image and performing iterative detail enhancement processing on the original image using at least one filtering function containing iteratively updated filtering parameters. Specifically, this includes: extracting feature information of the input image for the detail enhancement processing using at least one filtering function containing filtering parameters updated in the i-th iteration; and performing detail enhancement processing on the input image for the i-th iteration using the feature information from the i-th iteration to obtain a detail-enhanced image for the i-th iteration. The input image includes images related to the original image, and i is an integer greater than or equal to 1.

[0412] When the difference between the quality evaluation scores of the input image and the detail-enhanced image of the i-th iteration satisfies the iteration condition, the filtering parameters of the i-th iteration are updated to obtain the filtering parameters of the (i+1)-th iteration. The input image is then further enhanced with at least one filtering function containing the filtering parameters of the (i+1)-th iteration until the difference between the quality evaluation scores of the input image and the detail-enhanced image of the (i+1)-th iteration satisfies the iteration condition.

[0413] Shows an enhanced image of the last iteration.

[0414] Those skilled in the art will understand that embodiments of this disclosure can be provided as methods, systems, or computer program products. Therefore, this disclosure can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this disclosure can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage and optical storage) containing computer-usable program code.

[0415] This disclosure is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this disclosure. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create a machine for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 Devices that specify the functions in one or more boxes.

[0416] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including an instruction device, which is implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0417] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0418] Obviously, those skilled in the art can make various modifications and variations to this disclosure without departing from its spirit and scope. Therefore, if such modifications and variations fall within the scope of the claims of this disclosure and their equivalents, this disclosure is also intended to include such modifications and variations.

Claims

1. A display device system, wherein, include: Memory on which computer-readable instructions are stored; When a processor is configured to execute the computer-readable instructions, it performs the following steps: Acquire the original image, and perform iterative detail enhancement processing on the original image using at least one filtering function containing iteratively updated filtering parameters, specifically including: For the input image used in the detail enhancement process, at least one filtering function containing filtering parameters updated in the i-th iteration is used to extract the feature information of the input image in the i-th iteration. The input image is then subjected to detail enhancement processing in the i-th iteration using the feature information of the i-th iteration, resulting in a detail-enhanced image in the i-th iteration. The input image includes an image related to the original image, and i is an integer greater than or equal to 1. Specifically, the processor is configured to: extract the first feature information of the original image in the i-th iteration using a first filtering function containing first filtering parameters updated in the i-th iteration; perform detail enhancement processing on the original image in the i-th iteration using the first feature information of the i-th iteration, resulting in a first enhanced image in the i-th iteration; and extract the second feature information of the first enhanced image in the i-th iteration using a second filtering function containing second filtering parameters updated in the i-th iteration; perform detail enhancement processing on the first enhanced image in the i-th iteration using the second feature information of the i-th iteration, resulting in the detail-enhanced image in the i-th iteration. When the difference between the quality evaluation scores of the input image and the detail-enhanced image of the i-th iteration satisfies the iteration condition, the filtering parameters of the i-th iteration are updated to obtain the filtering parameters of the (i+1)-th iteration. The input image is then further enhanced with at least one filtering function containing the filtering parameters of the (i+1)-th iteration until the difference between the quality evaluation scores of the input image and the detail-enhanced image of the (i+1)-th iteration satisfies the iteration condition. The display device system also includes a display screen configured to display a detail-enhanced image of the last iteration.

2. The display device system according to claim 1, wherein, After acquiring the original image, before performing iterative detail enhancement processing on the original image using at least one filtering function containing iteratively updated filtering parameters, the processor is specifically configured to execute: When the original image is determined to be an RGB image, the original RGB image is converted to HSV color channels to obtain the original HSV image; The brightness V channel image of the original HSV image is used as the input image for the detail enhancement process.

3. The display device system according to claim 2, wherein, Before displaying the image from the last iteration of the detail enhancement process, the processor is specifically configured to perform: Convert the image processed in the last iteration of detail enhancement to an RGB detail-enhanced image; The display screen is specifically configured to perform: Display the RGB detail-enhanced image.

4. The display device system according to claim 1, wherein, The processor is specifically configured to execute: The first feature information of the i-th iteration is superimposed on the original image to obtain the first enhanced image of the i-th iteration; The second feature information of the i-th iteration is superimposed on the first enhanced image of the i-th iteration to obtain the detail enhanced image of the i-th iteration.

5. The display device system according to claim 1, wherein, The processor is specifically configured to execute: The original image is filtered using a first filtering function that includes the first filtering parameters updated in the i-th iteration to obtain the first filtered image in the i-th iteration; the original image is subtracted from the first filtered image in the i-th iteration to obtain the first feature information in the i-th iteration. The first enhanced image of the i-th iteration is filtered using a second filtering function that includes the second filtering parameters updated in the i-th iteration to obtain the second filtered image of the i-th iteration; the first enhanced image of the i-th iteration is subtracted from the second filtered image of the i-th iteration to obtain the second feature information of the i-th iteration.

6. The display device system according to claim 1, wherein, The first filtering function is a guided filtering function, and the second filtering function is a Gaussian filtering function; or, The first filtering function is a Gaussian filtering function, and the second filtering function is a guided filtering function.

7. The display device system according to claim 1, wherein, The quality assessment score of the original image is the first NIQE score obtained by evaluating the quality of the original image using the No Reference Image Quality Assessment (NIQE) algorithm. The quality evaluation score of the detail-enhanced image in the i-th iteration is the second NIQE score obtained by evaluating the quality of the detail-enhanced image in the i-th iteration using the NIQE algorithm.

8. The display device system according to claim 7, wherein, The processor is specifically configured to execute: The difference between the first NIQE score and the second NIQE score is determined to be less than or equal to zero or greater than or equal to a preset threshold.

9. The display device system according to any one of claims 1 to 8, wherein, After acquiring the original image, before performing iterative detail enhancement processing on the original image using at least one filtering function containing iteratively updated filtering parameters, the processor is specifically configured to execute: Using a super-resolution algorithm, the resolution of the input image processed by super-resolution is reconstructed to a preset resolution to obtain a reconstructed image, wherein the preset resolution is higher than the resolution of the original image; The input image for the super-resolution processing is either the original image or the contrast-enhanced image, wherein the contrast-enhanced image is obtained by performing contrast enhancement processing on the original image. The reconstructed image is used as the input image for the detail enhancement process.

10. The display device system according to claim 9, wherein, The processor is specifically configured to execute: The super-resolution factor corresponding to the size is determined based on the size of the original image; Based on the super-resolution factor and the super-resolution algorithm, the resolution of the original image is reconstructed to a preset resolution.

11. The display device system according to claim 9, wherein, After acquiring the original image, before performing iterative detail enhancement processing on the original image using at least one filtering function containing iteratively updated filtering parameters, the processor is specifically configured to execute: The reconstructed image after super-resolution processing or the original image without super-resolution processing is used as the input image, and iterative contrast enhancement processing is performed using the iteratively updated gamma value: The input image for the contrast enhancement process is subjected to gamma transformation according to the gamma value updated in the j-th iteration to obtain the contrast-enhanced image in the j-th iteration; where j is an integer greater than or equal to 1. If the contrast difference between the contrast-enhanced image of the j-th iteration and the contrast-enhanced input image is greater than the contrast threshold, the contrast of the contrast-enhanced input image is replaced with the contrast of the contrast-enhanced image of the j-th iteration, and the gamma value of the j-th iteration is updated to obtain the gamma value of the (j+1)-th iteration. Using the gamma value of the (j+1)-th iteration, the contrast-enhanced input image is further subjected to iterative contrast enhancement processing until the contrast difference between the contrast-enhanced image of the j-th iteration and the contrast-enhanced input image is less than or equal to the contrast threshold. The contrast-enhanced image from the last iteration is used as the input image for the detail enhancement process.

12. The display device system according to claim 11, wherein, Before using the contrast-enhanced image from the last iteration as the input image for the detail enhancement process, the processor is specifically configured to perform: Determine the difference in the mean brightness between the input image of the contrast enhancement process and the contrast enhancement image of the last iteration; The difference is superimposed with the brightness of the contrast-enhanced image from the last iteration to obtain a superimposed image; The overlaid image is used as the input image for the detail enhancement process.

13. The display device system according to claim 12, wherein, Prior to the iterative contrast enhancement process using the iteratively updated gamma values, the processor is specifically configured to perform: When it is determined that the reconstructed image after super-resolution processing is an RGB image, the reconstructed image is converted to HSV to obtain the original HSV image, and the V channel image of the original HSV image is used as the input image for the contrast enhancement processing. or, When the original image that has not undergone super-resolution processing is determined to be an RGB image, the original image is converted to HSV to obtain an original HSV image, and the V channel image of the original HSV image is used as the input image for the contrast enhancement processing.

14. The display device system according to claim 13, wherein, Before using the overlaid image as the input image for the detail enhancement processing, the processor is further configured to perform: The overlay image is converted into an RGB overlay image.

15. An adaptive image quality enhancement method, wherein, The method includes: The process involves acquiring an original image and performing iterative detail enhancement processing on it using at least one filtering function containing iteratively updated filtering parameters. Specifically, this includes: extracting feature information from the input image of the detail enhancement process in the i-th iteration using at least one filtering function containing filtering parameters updated in the i-th iteration; and performing iterative detail enhancement processing on the input image using the feature information updated in the i-th iteration to obtain a detail-enhanced image in the i-th iteration. The input image includes images related to the original image, and i is an integer greater than or equal to 1. The step of extracting feature information from the original image in the i-th iteration using at least one filtering function containing filtering parameters updated in the i-th iteration, and performing iterative detail enhancement processing on the input image using the feature information updated in the i-th iteration, yields a detail-enhanced image in the i-th iteration. The process of performing detail enhancement processing on the original image using iterative feature information to obtain a detail-enhanced image for the i-th iteration specifically includes: extracting the first feature information of the original image for the i-th iteration using a first filtering function containing the first filtering parameters updated for the i-th iteration; performing detail enhancement processing on the original image using the first feature information of the i-th iteration to obtain a first enhanced image for the i-th iteration; and extracting the second feature information of the first enhanced image for the i-th iteration using a second filtering function containing the second filtering parameters updated for the i-th iteration; performing detail enhancement processing on the first enhanced image using the second feature information of the i-th iteration to obtain the detail-enhanced image for the i-th iteration. When the difference between the quality evaluation scores of the input image and the detail-enhanced image of the i-th iteration satisfies the iteration condition, the filtering parameters of the i-th iteration are updated to obtain the filtering parameters of the (i+1)-th iteration. The input image is then further enhanced with at least one filtering function containing the filtering parameters of the (i+1)-th iteration until the difference between the quality evaluation scores of the input image and the detail-enhanced image of the (i+1)-th iteration satisfies the iteration condition. Shows an enhanced image of the last iteration.

16. The method according to claim 15, wherein, After acquiring the original image, and before performing iterative detail enhancement processing on the original image using at least one filtering function containing iteratively updated filtering parameters, the method further includes: When the original image is determined to be an RGB image, the original RGB image is converted to HSV color channels to obtain the original HSV image; The brightness V channel image of the original HSV image is used as the input image for the detail enhancement process.

17. The method according to claim 15, wherein, The first filtering function is a guided filtering function, and the second filtering function is a Gaussian filtering function; or, The first filtering function is a Gaussian filtering function, and the second filtering function is a guided filtering function.

18. The method according to claim 15, wherein, The quality assessment score of the original image is the first NIQE score obtained by evaluating the quality of the original image using the No Reference Image Quality Assessment (NIQE) algorithm. The quality evaluation score of the detail-enhanced image in the i-th iteration is the second NIQE score obtained by evaluating the quality of the detail-enhanced image in the i-th iteration using the NIQE algorithm.

19. The method according to any one of claims 15 to 18, wherein, After acquiring the original image, and before performing iterative detail enhancement processing on the original image using at least one filtering function containing iteratively updated filtering parameters, the method further includes: Using a super-resolution algorithm, the resolution of the input image for super-resolution processing is reconstructed to a preset resolution to obtain a reconstructed image, wherein the preset resolution is higher than the resolution of the original image; wherein the input image for super-resolution processing is either the original image or a contrast-enhanced image, and the contrast-enhanced image is obtained by performing contrast enhancement processing on the original image; The reconstructed image is used as the input image for the detail enhancement process.

20. The method according to claim 19, wherein, After acquiring the original image, and before performing iterative detail enhancement processing on the original image using at least one filtering function containing iteratively updated filtering parameters, the method further includes: The reconstructed image after super-resolution processing or the original image without super-resolution processing is used as the input image, and iterative contrast enhancement processing is performed using the iteratively updated gamma value: The input image for the contrast enhancement process is subjected to gamma transformation according to the gamma value updated in the j-th iteration to obtain the contrast-enhanced image in the j-th iteration; where j is an integer greater than or equal to 1. If the contrast difference between the contrast-enhanced image of the j-th iteration and the contrast-enhanced input image is greater than the contrast threshold, the contrast of the contrast-enhanced input image is replaced with the contrast of the contrast-enhanced image of the j-th iteration, and the gamma value updated in the j-th iteration is updated to obtain the gamma value of the (j+1)-th iteration. Using the gamma value of the (j+1)-th iteration, the contrast-enhanced input image is further subjected to iterative contrast enhancement processing until the contrast difference between the contrast-enhanced image of the j-th iteration and the contrast-enhanced input image is less than or equal to the contrast threshold. The contrast-enhanced image from the last iteration is used as the input image for the detail enhancement process.

21. A non-transient computer storage medium having a computer program stored thereon, wherein, When the program is executed by the processor, it implements the steps of the method as described in any one of claims 15 to 20.

Images