Image processing method, electronic equipment and computer readable medium
By decomposing the image into sub-images with different resolutions for denoising filtering and fusion processing, the problem of rough and loss of details of image noise removal in the prior art is solved, and a more refined noise removal effect and image details are achieved.
Patent Information
- Application Number
- CN202310942677.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2023-07-28
- Publication Date
- 2025-02-25
Smart Images
Figure CN119515707A_ABST
Abstract
Description
Technical Field
[0001] The present disclosure relates to the technical field of image processing, and in particular to an image processing method, an electronic device, and a computer-readable medium. Background Art
[0002] Image noise is the random variation of image information or pixel brightness, which is the degradation of image signals caused by some factors. Removing image noise during image processing is of great significance for improving the performance of shooting, monitoring, computer vision, etc.
[0003] When removing image noise, especially color noise, the filtering results are usually rough, with a significant loss of color saturation, and it is difficult to maintain image texture details. Summary of the invention
[0004] The embodiments of the present disclosure provide an image processing method, an electronic device, and a computer-readable medium.
[0005] In a first aspect, an embodiment of the present disclosure provides an image processing method, comprising: converting an image to be processed into multiple sub-images to be processed, wherein the resolutions of the multiple sub-images to be processed are different from each other; performing corresponding denoising filtering processing on the multiple sub-images to be processed respectively to obtain multiple filtered sub-images; and fusing the multiple filtered sub-images to obtain a target image.
[0006] In a second aspect, an embodiment of the present disclosure provides an electronic device, comprising: one or more processors; a memory on which one or more programs are stored, and when the one or more programs are executed by the one or more processors, the one or more processors implement the image processing method described in the first aspect of the embodiment of the present disclosure.
[0007] In a third aspect, an embodiment of the present disclosure provides a computer-readable medium having a computer program stored thereon, and when the program is executed by a processor, the image processing method described in the first aspect of the embodiment of the present disclosure is implemented.
[0008] In the image processing method provided by the embodiment of the present disclosure, the image to be processed is converted into a plurality of sub-images to be processed at different resolution scales, so that denoising filtering can be performed on the plurality of sub-images to be processed at different resolution scales respectively, and finally the plurality of filtered sub-images are fused to obtain a target image with noise removed, thereby achieving more precise control over the denoising process at different resolution scales, which is beneficial to improving the effect of noise removal. BRIEF DESCRIPTION OF THE DRAWINGS
[0009] Figure 1 is a flow chart of an image processing method in an embodiment of the present disclosure;
[0010] Figure 2 is a schematic diagram of a denoising filter process in an embodiment of the present disclosure;
[0011] Figure 3 is a schematic diagram of a bilateral filtering in an embodiment of the present disclosure;
[0012] Figure 4 is a schematic diagram of a fusion process in an embodiment of the present disclosure;
[0013] Figure 5 is a schematic diagram of an algorithm model for image processing in an embodiment of the present disclosure;
[0014] Figure 6 is a structural schematic diagram of an electronic device in an embodiment of the present disclosure;
[0015] Figure 7 It is a structural schematic diagram of a computer-readable medium in an embodiment of the present disclosure. DETAILED DESCRIPTION
[0016] In order to enable those skilled in the art to better understand the technical solution of the present disclosure, the image processing method, electronic device, and computer-readable medium provided by the present disclosure are described in detail below in conjunction with the accompanying drawings.
[0017] Example embodiments will be described more fully below with reference to the accompanying drawings, but the example embodiments may be embodied in different forms and should not be construed as limited to the embodiments set forth herein. On the contrary, the purpose of providing these embodiments is to make the present disclosure thorough and complete and to enable those skilled in the art to fully understand the scope of the present disclosure.
[0018] In the absence of conflict, the various embodiments of the present disclosure and the various features therein may be combined with each other.
[0019] As used herein, the term "and / or" includes any and all combinations of one or more of the associated listed items.
[0020] The terms used herein are only used to describe specific embodiments and are not intended to limit the present disclosure. As used herein, the singular forms "a", "an" and "the" are also intended to include the plural forms, unless the context clearly indicates otherwise. It will also be understood that when the terms "comprising" and / or "made of" are used in this specification, the presence of the features, wholes, steps, operations, elements and / or components is specified, but the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or groups thereof is not excluded.
[0021] Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by those of ordinary skill in the art. It will also be understood that terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant art and the present disclosure, and will not be interpreted as having an idealized or overly formal meaning unless explicitly defined as such herein.
[0022] First, refer to Figure 1 , the present disclosure provides an image processing method, including:
[0023] S1, converting the image to be processed into a plurality of sub-images to be processed, wherein the resolutions of the plurality of sub-images to be processed are different from each other;
[0024] S2, performing corresponding denoising filtering processing on the multiple sub-images to be processed respectively to obtain multiple filtered sub-images;
[0025] S3. Fusing the multiple filtered sub-images to obtain a target image.
[0026] The embodiments of the present disclosure do not impose any special restrictions on the image to be processed. In some embodiments, the image to be processed is a color image in a YUV image space.
[0027] In the embodiment of the present disclosure, the resolutions of the multiple sub-images to be processed are different from each other, that is, the image to be processed is decomposed at multiple resolution scales. The embodiment of the present disclosure does not specifically limit the number of sub-images to be processed and the resolution of each sub-image to be processed. In some embodiments, a pyramid can be constructed at any sampling rate, and the image to be processed can be decomposed at multiple resolution scales based on the pyramid.
[0028] In the disclosed embodiment, the image to be processed is decomposed into multiple sub-images to be processed at different resolution scales, so that denoising filtering can be performed on the multiple sub-images to be processed at different resolution scales respectively, and finally the multiple filtered sub-images are fused to obtain a target image with noise removed, thereby achieving more precise control of the denoising process at different resolution scales, which is beneficial to improving the effect of noise removal.
[0029] In some embodiments, when performing denoising filtering on a plurality of sub-images to be processed, denoising parameters can be controlled separately at different resolution scales to achieve fine control over the denoising process.
[0030] In the disclosed embodiment, according to the sub-images to be processed of different resolution scales, noises of different frequency bands can be distinguished, which is conducive to selecting corresponding noise reduction parameters corresponding to image noises of different frequency bands and determining corresponding denoising strengths, thereby improving the effect of noise removal.
[0031] The image processing method provided in the embodiments of the present disclosure can be applied to video recording and photo previewing scenes of vehicle-mounted cameras, driving recorders, terminal cameras, surveillance cameras, etc. The embodiments of the present disclosure do not specifically limit this.
[0032] The image processing method provided in the embodiments of the present disclosure may be used as a hardware unit in an application specific integrated circuit (ASIC) chip such as an image signal processor (ISP), or may be used as a digital image processing algorithm running on hardware such as a digital signal processor (DSP), a graphics processor (GPU), or a central processing unit (CPU) in a camera application. The embodiments of the present disclosure do not specifically limit this.
[0033] The disclosed embodiment does not specifically limit how to convert the image to be processed into multiple sub-images to be processed.
[0034] In some embodiments, the image to be processed is converted into a plurality of sub-images to be processed by downsampling the image to be processed.
[0035] Image sampling refers to the process of processing the pixel values of the original image through a certain sampling technology to obtain result images of different resolutions; if the resolution of the output result image is smaller than the original image, it is called downsampling; if the resolution of the output result image is larger than the original image, it is called upsampling.
[0036] In the embodiment of the present disclosure, when the image to be processed is converted into a plurality of sub-images to be processed by downsampling, different sampling rates correspond to different resolutions of the sub-images to be processed.
[0037] Accordingly, in some embodiments, converting the image to be processed into a plurality of sub-images to be processed includes:
[0038] The image to be processed is downsampled a preset number of times according to at least one sampling rate to obtain a plurality of sub-images to be processed.
[0039] In some embodiments, the image to be processed is downsampled according to a plurality of sampling rates to obtain sub-images to be processed with different resolutions.
[0040] In some embodiments, when downsampling the image to be processed according to a sampling rate, the image to be processed is downsampled in multiple layers, and each layer of downsampling obtains a sub-image to be processed.
[0041] Accordingly, in some embodiments, downsampling the image to be processed a preset number of times according to at least one sampling rate to obtain a plurality of sub-images to be processed includes:
[0042] The image to be processed is subjected to multi-layer downsampling according to a preset sampling rate to obtain the sub-image to be processed corresponding to each layer of downsampling.
[0043] In some embodiments, downsampling the image to be processed a preset number of times according to at least one sampling rate to obtain a plurality of sub-images to be processed includes:
[0044] According to the sampling rate S i Downsample the i-th sub-image to be downsampled to obtain the i+1-th sub-image to be processed;
[0045] Wherein, i is a positive integer and i≤n, n is a positive integer, n represents the preset number of times, and the first sub-image to be processed is the image to be processed.
[0046] In some embodiments, the plurality of sub-images to be processed are respectively subjected to corresponding denoising filtering processes to obtain a plurality of filtered sub-images, including:
[0047] Perform denoising filtering on the j-th sub-image to be processed to obtain the j-th filtered sub-image; wherein j is a positive integer and j≤n+1.
[0048] In some embodiments, the plurality of filtered sub-images are fused to obtain a target image, including:
[0049] According to the sampling rate S i Upsample the i+1th fused sub-image to obtain the i-th upsampled sub-image;
[0050] Fusing the i-th up-sampled sub-image with the i-th filtered sub-image to obtain an i-th fused sub-image;
[0051] Using the first fused sub-image obtained as the target image;
[0052] When i=n, the i+1th fused sub-image is the i+1th filtered sub-image.
[0053] In some embodiments, fusing the i-th upsampled sub-image with the i-th filtered sub-image to obtain the i-th fused sub-image includes:
[0054] According to the texture feature of the i-th filtered sub-image, the i-th up-sampled sub-image is fused with the i-th filtered sub-image to obtain the i-th fused sub-image.
[0055] In some embodiments, according to the texture feature of the i-th filtered sub-image, the i-th up-sampled sub-image is fused with the i-th filtered sub-image to obtain the i-th fused sub-image, including:
[0056] Performing flat area detection on the i-th filtered sub-image to obtain a fusion coefficient;
[0057] According to the fusion coefficient, a weighted sum is performed on the i-th filtered sub-image and the i-th up-sampled sub-image to obtain the i-th fused sub-image.
[0058] In the embodiment of the present disclosure, performing multi-layer downsampling on the image to be processed means performing multiple downsampling on the image to be processed, wherein the latter downsampling is performed on the basis of the result image obtained by the previous downsampling. When performing multi-layer downsampling on the image to be processed, the resolution of the sub-image to be processed corresponding to each layer of downsampling is different, and the resolution of the sub-image to be processed corresponding to the latter layer of downsampling is smaller than the resolution of the sub-image to be processed corresponding to the previous layer of downsampling.
[0059] The embodiments of the present disclosure do not impose any special restrictions on the number of downsampling times when performing multi-layer downsampling on the image to be processed. If there is sufficient memory, any number of downsampling can be performed. In some embodiments, the size of the sub-image to be processed obtained by multi-layer downsampling is not less than the size of the filter window for performing denoising filtering.
[0060] In some embodiments, when implemented in hardware, the number of downsampling does not exceed 4 considering limitations such as buffer size.
[0061] In the embodiment of the present disclosure, when downsampling the image to be processed, integer multiple sampling or non-integer multiple sampling may be supported, and the embodiment of the present disclosure does not specifically limit this.
[0062] The disclosed embodiment does not specifically limit how to downsample the image to be processed.
[0063] In some embodiments, the image to be processed is downsampled using bilinear sampling in a post-pass.
[0064] For example, suppose the width and height of the original image are W1 and H1 respectively, and the width and height of the output image are W2 and H2 respectively. For each pixel point I_dst(x,y) of the output image, the sampling value of its pixel value is:
[0065] I_dst(x,y)=a*I(x1,y1)+b*I(x2,y1)+c*I(x1,y2)+d*f(x2,y2);
[0066] in,
[0067]
[0068] in, Indicates rounding down. Indicates rounding up.
[0069] The disclosed embodiment does not specifically limit how to perform denoising filtering on the sub-image to be processed.
[0070] In some embodiments, performing denoising filtering on the plurality of sub-images to be processed respectively includes:
[0071] Generate a guidance map according to the sub-image to be processed;
[0072] A denoising filtering process is performed on the sub-image to be processed according to the guide image to obtain a filtered sub-image corresponding to the sub-image to be processed.
[0073] In the embodiment of the present disclosure, the guide map may be a texture map, a brightness map, a color saturation map, etc., and the embodiment of the present disclosure does not specifically limit this.
[0074] In the disclosed embodiment, a denoising filter is performed on the sub-image to be processed using a guide map, and a suitable filter strength can be selected for the filter area, thereby neither losing image details due to excessive filter strength nor causing residual noise due to insufficient filter strength, which is beneficial to improving the denoising filter effect.
[0075] In some embodiments, the denoising filtering process using the guide map can be implemented as follows: Figure 2 The filter unit shown.
[0076] exist Figure 2 In the method, the filtering unit generates a guiding map guide_map according to the input image img_in through the guiding map generating unit GeneratingGuiding Map, and the guided filtering unit Guided Filtering performs denoising filtering processing according to the input image img_in and the guiding map guide_map to generate a filtered image.
[0077] In some embodiments, performing denoising filtering on the sub-image to be processed according to the guide map includes:
[0078] Determining the filtering strength of the sub-image to be processed according to the guide map;
[0079] The sub-image to be processed is subjected to denoising filtering according to the filtering strength to obtain the corresponding filtered sub-image.
[0080] In the embodiment of the present disclosure, the denoising filtering process may be performed based on Gaussian filtering, median filtering, bilateral filtering, etc. The embodiment of the present disclosure does not specifically limit this.
[0081] In some embodiments, the process of bilateral filtering based on the color saturation guide map is as follows: Figure 3 shown.
[0082] exist Figure 3 In the processing shown, since color noise usually appears in low saturation areas, and the greater the color noise intensity, the greater the bilateral filtering intensity required, and excessive filtering intensity in high saturation areas will lose image saturation. In some embodiments, the color distance parameter in bilateral filtering is adaptively adjusted in the following manner:
[0083]
[0084] Among them, σ′ color is the adjusted bilateral filter color distance parameter, σ color is the color distance parameter in the bilateral filtering before adjustment, α is the saturation factor, and β=1-α.
[0085] The disclosed embodiment does not impose any special limitation on how to fuse multiple filtered sub-images.
[0086] In some embodiments, a plurality of the filtered sub-images are fused to obtain a target image, including:
[0087] According to the texture features of the filtered sub-images, a plurality of the filtered sub-images are fused to obtain the target image.
[0088] In the disclosed embodiment, fusion processing is performed based on texture features, which is beneficial for preserving more high-frequency texture details when selecting a region for more adequate denoising.
[0089] In some embodiments, the up-sampled sub-image is fused with the filtered sub-image according to the texture feature of the filtered sub-image, including:
[0090] Performing flat area detection on the filtered sub-image to obtain a fusion coefficient;
[0091] generating an upsampled sub-image corresponding to the filtered sub-image;
[0092] According to the fusion coefficient, a weighted sum is performed on the filtered sub-image and the up-sampled sub-image.
[0093] In some embodiments, the fusion processing based on texture features can be implemented as follows: Figure 4 The spatial fusion unit shown.
[0094] exist Figure 4 In the figure, there are two images input to the spatial fusion unit, one is the upsampled image img_up, and the other is the filtered image img_blur. The flat area detection unit Flat Area Mesuring performs flat area detection on the image img_blur, and outputs a fusion coefficient α ranging from 0 to 1. Then, the image img_up and the image img_blur are weighted summed according to the following formula to obtain the fused image img_rec:
[0095] img_rec=(1-α)*img_blur+α*img_up.
[0096] The embodiment of the present disclosure does not impose any special limitation on the specific algorithm implementation of steps S1 to S3.
[0097] In some embodiments, steps S1 to S3 are implemented as follows Figure 5 The algorithm model shown.
[0098] exist Figure 5 In the example, img_floor_0 is the image to be processed transmitted from the system bus; Filtering Uint is the denoising unit, also known as the filtering unit; Down Sampling Uint is the downsampling unit; Up Sampling Uint is the upsampling unit; Blending Uint is the fusion unit. After three downsamplings, img_floor_0 outputs three sub-images to be processed with different resolution scales: img_floor_1, img_floor_2, and img_floor_3. These sub-images to be processed with different resolutions then enter the denoising unit, and after filtering, they output filtered sub-images: img_floor_0_blur, img_floor_1_blur, img_floor_2_blur, and img_floor_3_blur. Then, upsampling starts from the bottom layer. After passing through the upsampling unit, the filtered sub-image of the same layer and the upsampled sub-image of the corresponding layer: img_floor_2_up, img_floor_1_up, img_floor_0_up, enter the fusion unit, and output the fused sub-images: img_floor_2_rec, img_floor_1_rec, img_floor_0_rec. Finally, the fused image img_floor_0_rec is the target image, that is, the final denoising result image; return to the bus and enter the next ISP module.
[0099] Corresponds to Figure 5In some embodiments, the algorithm model shown in the figure performs multi-layer downsampling on the image to be processed according to a preset sampling rate to obtain the sub-image to be processed corresponding to each layer of downsampling, including:
[0100] Down-sampling the image to be processed according to the preset sampling rate to obtain a first sub-image to be processed;
[0101] Down-sampling the first sub-image to be processed according to the preset sampling rate to obtain a second sub-image to be processed;
[0102] The second sub-image to be processed is down-sampled according to the preset sampling rate to obtain a third sub-image to be processed.
[0103] Corresponds to Figure 5 In some embodiments, the algorithm model shown in FIG. 1 performs denoising filtering on a plurality of sub-images to be processed to obtain a plurality of filtered sub-images, including:
[0104] Performing denoising filtering on the first sub-image to be processed to obtain a first filtered sub-image;
[0105] Performing denoising filtering on the second sub-image to be processed to obtain a second filtered sub-image;
[0106] Performing denoising filtering on the third sub-image to be processed to obtain a third filtered sub-image;
[0107] Perform denoising filtering on the image to be processed to obtain a fourth filtered sub-image.
[0108] Corresponds to Figure 5 In some embodiments, the algorithm model shown in FIG. 1 is used to fuse a plurality of the filtered sub-images to obtain a target image, including:
[0109] Upsampling the third filtered sub-image to obtain a third up-sampled sub-image;
[0110] Fusing the third up-sampled sub-image with the second filtered sub-image to obtain a second fused sub-image;
[0111] Upsampling the second fused sub-image to obtain a second upsampled sub-image;
[0112] The second up-sampled sub-image is fused with the first filtered sub-image to obtain a first fused sub-image; the first fused sub-image is up-sampled to obtain a first up-sampled sub-image
[0113] The first up-sampled sub-image is fused with the fourth filtered sub-image to obtain the target image.
[0114] The embodiments of the present disclosure have the following beneficial effects:
[0115] First, the multi-resolution scale method can more finely fuse information of different scales, achieve more refined control and more diverse effects, and use noise estimation to more finely control the multi-scale decomposition process;
[0116] Second, motion estimation uses a multi-scale recursive approach, which can limit the search space at each scale and improve computational efficiency. In addition, the edge map estimation method makes the estimation more robust.
[0117] Third, spatial filtering adopts an improved version of the non-local mean method, which can achieve better visual effects and reuse the search process of motion estimation to improve the efficiency of the entire algorithm;
[0118] Fourth, overall, this algorithm can achieve better results, faster speed, more precise control, and more diverse effects;
[0119] Second, refer to Figure 6 , an embodiment of the present disclosure provides an electronic device, comprising:
[0120] One or more processors 101;
[0121] A memory 102 storing one or more programs, which, when executed by one or more processors, enables the one or more processors to implement the image processing method according to the first aspect of the embodiment of the present disclosure;
[0122] One or more I / O interfaces 103 are connected between the processor and the memory and are configured to implement information exchange between the processor and the memory.
[0123] Among them, the processor 101 is a device with data processing capabilities, including but not limited to a central processing unit (CPU) and the like; the memory 102 is a device with data storage capabilities, including but not limited to random access memory (RAM, more specifically SDRAM, DDR, etc.), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory (FLASH); the I / O interface (read-write interface) 103 is connected between the processor 101 and the memory 102, and can realize information interaction between the processor 101 and the memory 102, including but not limited to a data bus (Bus) and the like.
[0124] In some embodiments, the processor 101 , the memory 102 , and the I / O interface 103 are connected to each other via a bus 104 , and further connected to other components of the computing device.
[0125] Thirdly, refer to Figure 7The embodiment of the present disclosure provides a computer-readable medium on which a computer program is stored. When the program is executed by a processor, the image processing method described in the first aspect of the embodiment of the present disclosure is implemented.
[0126] It will be appreciated by those skilled in the art that all or some of the steps, systems, and functional modules / units in the methods disclosed above may be implemented as software, firmware, hardware, and appropriate combinations thereof. In hardware implementations, the division between the functional modules / units mentioned in the above description does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or a function or step may be performed by several physical components in cooperation. Some or all physical components may be implemented as software executed by a processor, such as a central processing unit, a digital signal processor, or a microprocessor, or implemented as hardware, or implemented as an integrated circuit, such as an application-specific integrated circuit. Such software may be distributed on a computer-readable medium, which may include a computer storage medium (or non-transitory medium) and a communication medium (or temporary medium). As known to those skilled in the art, the term computer storage medium includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and can be accessed by a computer. In addition, it is well known to those skilled in the art that communication media typically contain computer-readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media.
[0127] Example embodiments have been disclosed herein, and although specific terms are employed, they are used and should be interpreted only in a general illustrative sense and not for limiting purposes. In some instances, it will be apparent to those skilled in the art that, unless otherwise expressly noted, features, characteristics, and / or elements described in conjunction with a particular embodiment may be used alone or in combination with features, characteristics, and / or elements described in conjunction with other embodiments. Therefore, those skilled in the art will appreciate that various changes in form and detail may be made without departing from the scope of the present disclosure as set forth in the appended claims.
Claims
1. An image processing method, comprising: Converting the image to be processed into a plurality of sub-images to be processed, wherein the resolutions of the plurality of sub-images to be processed are different from each other; Performing corresponding denoising filtering processing on the multiple sub-images to be processed respectively to obtain multiple filtered sub-images; The multiple filtered sub-images are fused to obtain a target image.
2. The image processing method according to claim 1, wherein: Convert the image to be processed into multiple sub-images to be processed, including: The image to be processed is downsampled a preset number of times according to at least one sampling rate to obtain a plurality of sub-images to be processed.
3. The image processing method according to claim 2, wherein: Downsampling the image to be processed a preset number of times according to at least one sampling rate to obtain a plurality of sub-images to be processed includes: According to the sampling rate S i Downsample the i-th sub-image to be downsampled to obtain the i+1-th sub-image to be processed; Wherein, i is a positive integer and i≤n, n is a positive integer, n represents the preset number of times, and the first sub-image to be processed is the image to be processed.
4. The image processing method according to claim 3, wherein: Perform corresponding denoising filtering processing on the multiple sub-images to be processed respectively to obtain multiple filtered sub-images, including: Perform denoising filtering on the j-th sub-image to be processed to obtain the j-th filtered sub-image; wherein j is a positive integer and j≤n+1.
5. The image processing method according to claim 4, wherein: The plurality of filtered sub-images are fused to obtain a target image, including: According to the sampling rate S i Upsample the i+1th fused sub-image to obtain the i-th upsampled sub-image; Fusing the i-th up-sampled sub-image with the i-th filtered sub-image to obtain an i-th fused sub-image; Using the first fused sub-image obtained as the target image; When i=n, the i+1th fused sub-image is the i+1th filtered sub-image.
6. The image processing method according to claim 5, wherein: The i-th up-sampled sub-image is fused with the i-th filtered sub-image to obtain the i-th fused sub-image, comprising: According to the texture feature of the i-th filtered sub-image, the i-th up-sampled sub-image is fused with the i-th filtered sub-image to obtain the i-th fused sub-image.
7. The image processing method according to claim 6, wherein: According to the texture feature of the i-th filtered sub-image, the i-th up-sampled sub-image is fused with the i-th filtered sub-image to obtain the i-th fused sub-image, including: Performing flat area detection on the i-th filtered sub-image to obtain a fusion coefficient; According to the fusion coefficient, a weighted sum is performed on the i-th filtered sub-image and the i-th up-sampled sub-image to obtain the i-th fused sub-image.
8. The image processing method according to any one of claims 1 to 7, wherein: Performing corresponding denoising filtering processing on the multiple sub-images to be processed respectively, including for each of the sub-images to be processed: Generate a guidance map according to the sub-image to be processed; A denoising filtering process is performed on the sub-image to be processed according to the guide image to obtain a filtered sub-image corresponding to the sub-image to be processed.
9. The image processing method according to claim 8, wherein: Performing denoising filtering on the sub-image to be processed according to the guide map, comprising: Determining the filtering strength of the sub-image to be processed according to the guide map; The sub-image to be processed is subjected to denoising filtering according to the filtering strength to obtain the corresponding filtered sub-image.
10. An electronic device comprising: one or more processors; A memory having one or more programs stored thereon, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the image processing method according to any one of claims 1 to 9.
11. A computer-readable medium having a computer program stored thereon, wherein when the program is executed by a processor, the image processing method according to any one of claims 1 to 9 is implemented.