Image denoising method, system and readable storage medium

Abstract

This invention relates to an image denoising method, system, and readable storage medium. The method includes: performing Gaussian filtering and downsampling on a current frame image to obtain a first downsampled image; performing Gaussian pyramid decomposition to obtain a current frame image group; upsampling to obtain an upsampled image, and subtracting it from the current frame image to obtain high-frequency information; performing motion estimation and texture estimation at different scales on the current frame image group and a reference frame image group, and correcting them with guided filtering; performing spatiotemporal filtering on the first downsampled image and the reference frame image based on motion estimation weights; upsampling the denoised downsampled image to obtain a denoised upsampled image; and weighted fusing the denoised upsampled image and high-frequency information according to texture estimation weights to obtain a final image; downsampling the final image and performing Gaussian pyramid decomposition to obtain a reference frame image group for denoising the next frame image. This invention can preserve more texture details while increasing the image signal-to-noise ratio.

Description

Technical Field

[0001] This invention relates to the field of image processing technology, and in particular to an image noise reduction method, system, and readable storage medium. Background Technology

[0002] Cameras using CMOS sensors are inevitably affected by noise during video recording, storage, and transmission, leading to reduced image quality. Image denoising removes noise from the current signal to improve the image signal-to-noise ratio and restore the information of the real scene. For video sequences, this can be viewed as denoising multiple frames. Since there is correlation between different frames, temporal denoising is also used to improve the image signal-to-noise ratio. Generally, video denoising can be divided into temporal denoising and spatial denoising.

[0003] Spatial domain denoising algorithms are generally categorized into filter-based, local similarity-based, transform domain-based, and data-driven methods. Filter-based methods typically use low-pass filters to filter the image, removing high-frequency noise to obtain a less noisy image, such as Gaussian filters and mean filters. Local similarity-based methods leverage the similarity of textures in the image's neighborhood, using weighted averaging to improve the signal-to-noise ratio (SNR). Transform domain-based methods convert the image signal to other signal domains, such as frequency domains, and filter out noise using high-frequency characteristics before converting back to the spatial domain, resulting in an image with a higher SNR. Data-driven methods, represented by deep learning, use large amounts of data to train neural networks to fit a noise representation, then perform forward inference on the input online to obtain either a noisy image or a clean image with noise removed, thereby improving the SNR.

[0004] Temporal denoising algorithms are generally implemented by superimposing multiple frames. For example, IIR filters are used to average multiple past frames to achieve a better signal-to-noise ratio. Alternatively, frames before and after the current frame are cached to predict the motion vector of the current frame image by leveraging the continuity, thereby assisting in better temporal denoising.

[0005] Generally, temporal or spatial denoising alone is insufficient to completely suppress noise in video data. Therefore, a combination of temporal and spatial denoising is necessary to improve the signal-to-noise ratio while maintaining image resolution. However, this combined temporal and spatial denoising method requires caching at least one full frame of reference image to achieve temporal denoising. In an era of continuously increasing video resolution, the bandwidth pressure on actual hardware devices is increasing daily, posing a serious challenge to the power consumption performance of the devices. At the same time, the accuracy of motion estimation also affects the performance of temporal and spatial denoising, making it difficult to filter out some weak texture areas with poor signal-to-noise ratios, resulting in the loss of texture details. Summary of the Invention

[0006] The purpose of this invention is to provide an image denoising method, system, and readable storage medium, which aims to reduce the bandwidth required for reference frame image buffering, while improving the image signal-to-noise ratio and removing large-particle noise without reducing image resolution.

[0007] To achieve the above objectives, the present invention provides an image denoising method, comprising:

[0008] The current frame image is subjected to Gaussian filtering and first downsampling to obtain the first downsampled image. The first downsampled image is then subjected to Gaussian pyramid decomposition to obtain a current frame image group including multiple scale images.

[0009] The first downsampled image is upsampled to obtain an upsampled image, and the high-frequency information of the current frame image is obtained by subtracting the current frame image from the upsampled image.

[0010] Motion and texture estimation are performed on the current frame image group and the pre-cached reference frame image group at different scales, and the motion estimation weights and texture estimation weights at each scale are corrected by guided filtering.

[0011] Based on the motion estimation weights at the corresponding scale, the first downsampled image and the reference frame images at the corresponding scale in the reference frame image group are subjected to spatiotemporal filtering to obtain the noise-reduced downsampled image;

[0012] The downsampled image after noise reduction is upsampled to obtain a denoised upsampled image. The upsampled image after noise reduction is then weighted and fused with the high-frequency information according to the texture estimation weights at the corresponding scale to obtain the final image.

[0013] The final image is downsampled and decomposed into a Gaussian pyramid to obtain a set of reference frame images for noise reduction of the next frame.

[0014] Optionally, the maximum size of the reference frame image group is the image size after the reference frame image has been downsampled for the first time.

[0015] Optionally, the motion estimation can be performed using inter-frame difference, optical flow, or density-based clustering algorithms.

[0016] Optionally, the texture estimation can be performed using an edge extraction operator.

[0017] Optionally, when correcting the motion estimation weights and texture estimation weights at each scale through guided filtering, the weight information is calculated and corrected layer by layer from the smallest scale to the largest scale.

[0018] Based on the same technical concept, the present invention also provides an image noise reduction system, comprising:

[0019] The first decomposition module is configured to perform Gaussian filtering on the current frame image and perform a first downsampling to obtain a first downsampled image, and perform Gaussian pyramid decomposition on the first downsampled image to obtain a current frame image group including multiple scale images.

[0020] The upsampling module is configured to upsample the first downsampled image to obtain an upsampled image, and to subtract the current frame image from the upsampled image to obtain the high-frequency information of the current frame image;

[0021] The estimation module is configured to perform motion estimation and texture estimation at different scales on the current frame image group and the pre-cached reference frame image group, and to correct the motion estimation weights and texture estimation weights at each scale through guided filtering;

[0022] The filtering module is configured to perform spatiotemporal filtering on the first downsampled image and the reference frame images of the corresponding scale in the reference frame image group based on the motion estimation weights at the corresponding scale, so as to obtain the noise-reduced downsampled image;

[0023] The fusion module is configured to upsample the denoised downsampled image to obtain a denoised upsampled image, and then weight and fuse the denoised upsampled image with the high-frequency information according to the texture estimation weights at the corresponding scale to obtain the final image.

[0024] The second decomposition module downsamples the final image and performs Gaussian pyramid decomposition to obtain a group of reference frame images for noise reduction of the next frame image.

[0025] Optionally, the maximum size of the reference frame image group is the image size after the reference frame image has been downsampled for the first time.

[0026] Optionally, the motion estimation can be performed using inter-frame difference, optical flow, or density-based clustering algorithms.

[0027] Optionally, the texture estimation can be performed using an edge extraction operator.

[0028] Based on the same technical concept, the present invention also provides a readable storage medium having a computer program stored thereon, which, when executed, can implement the image noise reduction method described above.

[0029] In the image denoising method, system, and readable storage medium provided by this invention, motion estimation and texture estimation are performed on multi-scale images, and the weights at each scale are corrected by guided filtering. More accurate motion estimation weights and texture estimation weights are obtained, which are then used for spatiotemporal filtering and fusion of high-frequency information. This increases the image signal-to-noise ratio while preserving more texture details, resulting in a denoising result with better signal-to-noise ratio and richer texture. Furthermore, the reference frame image group used for denoising the next frame is obtained by downsampling the current image and performing Gaussian pyramid decomposition, effectively saving transmission bandwidth. Attached Figure Description

[0030] Those skilled in the art will understand that the accompanying drawings are provided to better understand the invention and do not constitute any limitation on the scope of the invention. Wherein:

[0031] Figure 1 This is a flowchart of an image noise reduction method provided in an embodiment of the present invention. Detailed Implementation

[0032] To make the objectives, advantages, and features of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that the drawings are in a very simplified form and use non-precise proportions, and are only used to facilitate and clearly illustrate the purpose of the embodiments of this invention. Please refer to the accompanying drawings to make the objectives, features, and advantages of this invention more apparent and understandable. It should be understood that the structures, proportions, sizes, etc., depicted in the accompanying drawings are only used to complement the content disclosed in the specification, for those skilled in the art to understand and read, and are not intended to limit the implementation conditions of this invention. Any modifications to the structure, changes in proportions, or adjustments to the size, if they are the same as or similar to the effects and objectives achieved by this invention, should still fall within the scope of the technical content disclosed in this invention.

[0033] As used herein, the singular forms “a,” “an,” and “the” include plural objects unless otherwise expressly stated. As used herein, the term “or” is generally used to mean “and / or” unless otherwise expressly stated. As used herein, the term “a number” is generally used to mean “at least one” unless otherwise expressly stated. As used herein, the term “at least two” is generally used to mean “two or more” unless otherwise expressly stated.

[0034] Please refer to Figure 1 This invention provides an image noise reduction method, comprising the following steps:

[0035] S1. Perform Gaussian filtering on the current frame image and perform the first downsampling to obtain the first downsampled image. Perform Gaussian pyramid decomposition on the first downsampled image to obtain a current frame image group including multiple scale images.

[0036] S2. Upsample the first downsampled image to obtain an upsampled image, and subtract the current frame image from the upsampled image to obtain the high-frequency information of the current frame image;

[0037] S3. Perform motion estimation and texture estimation at different scales on the current frame image group and the pre-cached reference frame image group, and correct the motion estimation weights and texture estimation weights at each scale through guided filtering;

[0038] S4. Based on the motion estimation weights at the corresponding scale, the first downsampled image and the reference frame images at the corresponding scale in the reference frame image group are subjected to spatiotemporal filtering to obtain the noise-reduced downsampled image.

[0039] S5. Upsample the denoised downsampled image to obtain a denoised upsampled image. Then, weighted fuse the denoised upsampled image with the high-frequency information according to the texture estimation weights at the corresponding scale to obtain the final image.

[0040] S6. The final image is downsampled and decomposed into Gaussian pyramids to obtain a group of reference frame images for noise reduction of the next frame image.

[0041] This embodiment performs motion and texture estimation on multi-scale images and corrects the weights at each scale through guided filtering, resulting in more accurate motion and texture estimation weights. These weights are then used for spatiotemporal filtering and fusion of high-frequency information, increasing the image signal-to-noise ratio while preserving more texture details, thus achieving a denoising result with better signal-to-noise ratio and richer texture. Furthermore, the reference frame image group used for denoising the next frame is obtained by downsampling the current image and performing Gaussian pyramid decomposition, effectively saving transmission bandwidth.

[0042] First, S1 is executed, where Gaussian filtering is applied to the current frame image, followed by a first downsampling to obtain a first downsampled image. Gaussian pyramid decomposition is then performed on the first downsampled image to obtain a group of current frame images comprising multiple scales. Next, S2 is executed, where the first downsampled image is upsampled to obtain an upsampled image. The difference between the current frame image and the upsampled image is then calculated to obtain the high-frequency information of the current frame image.

[0043] Specifically, the current frame image is first acquired, denoted as Q, with a size of H x W. Then, the current frame image Q is downsampled to obtain Q1. Downsampling methods include, but are not limited to, average pooling, single-point selection, etc., with a size of H / 2 x W / 2. Next, Q1 is upsampled to obtain Q′. Upsampling methods include, but are not limited to, interpolation, padding, etc. The difference between Q and Q′ yields the high-frequency information H.

[0044] In this embodiment, the downsampled image Q1 is decomposed into k layers using a Gaussian pyramid. The decomposition process for the i-th layer (0≤i≤k) involves first using a pyramid with radius r0 and variance δ. 2 The image is low-pass filtered using a Gaussian kernel and then downsampled to obtain the (i+1)th layer with a resolution of 1 / 4 of the previous layer, expressed by the formula:

[0045]

[0046] Then, S3 is executed to perform motion estimation and texture estimation at different scales on the current frame image group and the pre-cached reference frame image group, and to correct the motion estimation weights and texture estimation weights at each scale through guided filtering.

[0047] In this embodiment, the pre-cached reference frame image group is obtained by downsampling the previous frame image and then performing Gaussian pyramid decomposition. Therefore, the maximum size of the reference frame image group is the image size after the first downsampling of the reference frame image. In the i-th layer, with Q... i A reference frame image of the same scale is denoted as P. i The "performing motion estimation and texture estimation at different scales" mentioned in this application refers to performing motion estimation and texture estimation on the current frame image and the reference frame image at the same scale, repeating this process multiple times to obtain motion estimation and texture estimation results for the current frame image and the reference frame image at various different scales. For example, in the i-th layer, for the current frame image Q... i and the reference frame image P i Motion estimation and texture estimation are performed to obtain motion estimation weights M at different scales. i and texture estimation weight D i .

[0048] Optionally, the motion estimation can be performed using inter-frame difference, optical flow, or density-based clustering algorithms.

[0049] Optionally, edge extraction operators such as the Sobel operator, Laplace operator, etc., can be used for the texture estimation.

[0050] Obtain motion estimation weights M at different scales i and texture estimation weight D iThen, guided filtering is used to correct the motion estimation weights and texture estimation weights at each scale. Specifically, taking the correction of motion estimation weights as an example, let the corrected motion estimation weights of the i-th layer be M′. i The corrected estimated weights M′ of the k-th layer at the minimum scale k =M k The calculation process for the corrected weights involves upsampling the corrected estimated weights of layer i+1 as a guide graph, and then applying guided filtering to the uncorrected estimated weights of layer i to obtain M′. i The motion weight correction is ultimately passed to layer 1, resulting in a corrected motion estimation weight of size H / 2 x W / 2, denoted as M1′. Similarly, the texture weight correction is passed to layer 0, resulting in a corrected texture estimation weight of size H x W, denoted as D0′.

[0051] Then, S4 is executed, where the first downsampled image and the reference frame images of the corresponding scale in the reference frame image group are subjected to spatiotemporal filtering based on the motion estimation weights at the corresponding scale to obtain the denoised downsampled image. Specifically, at the first scale, spatiotemporal filtering is performed on the current frame image at the same scale based on the modified motion estimation weights M1′, which is divided into spatial filtering and temporal filtering fusion.

[0052] In the first downsampled image Q i Spatial filtering is performed at the corresponding scale using nonlocal mean filtering. Within an NxN window centered on the current point, nxn block matching is performed, and the similarity between blocks is calculated. Then, the similarity is mapped to fusion weights for weighted fusion, as follows:

[0053] Let the coordinates of the center point of the central block be (x0, y0), and the coordinates of the center point of the block whose similarity is to be calculated be (x1, y1). Then the similarity can be expressed by the formula:

[0054]

[0055] Let the spatial filtering result be denoted as S, then the weighted fusion process can be expressed by the formula:

[0056]

[0057] Where k1 and k2 are hyperparameters, the temporal filtering is the weighted fusion of the spatial filtering result S of the current frame image and the reference frame image P1 at the corresponding scale according to the weight α, and the temporal fusion result is denoted as T, as shown in the following formula:

[0058] T=αS+(1-α)P1

[0059]

[0060] k3 and k4 are hyperparameters.

[0061] Next, step S5 is executed to upsample the denoised downsampled image to obtain a denoised upsampled image. This denoised upsampled image has the same size as the current frame image. Based on the texture estimation weights at the corresponding scale, the denoised upsampled image and the high-frequency information are weighted and fused to obtain the final image. In this embodiment, the temporal fusion result T is upsampled to obtain T'. At the 0th scale, based on the texture weights W... D The high-frequency information H obtained by fusing S2 is denoted as O in the final spatiotemporal domain denoised image, and can be expressed by the formula:

[0062]

[0063] O = T' + W D H

[0064] k5 and k6 are hyperparameters.

[0065] Finally, S6 is executed to downsample the final image and perform Gaussian pyramid decomposition to obtain a reference frame image group for noise reduction of the next frame image, in order to save bandwidth consumption.

[0066] Based on the same inventive concept, this invention also proposes an image denoising system for implementing the image denoising method described above, comprising:

[0067] The first decomposition module is configured to perform Gaussian filtering on the current frame image and perform a first downsampling to obtain a first downsampled image, and perform Gaussian pyramid decomposition on the first downsampled image to obtain a current frame image group including multiple scale images.

[0068] The upsampling module is configured to upsample the first downsampled image to obtain an upsampled image, and to subtract the current frame image from the upsampled image to obtain the high-frequency information of the current frame image;

[0069] The estimation module is configured to perform motion estimation and texture estimation at different scales on the current frame image group and the pre-cached reference frame image group, and to correct the motion estimation weights and texture estimation weights at each scale through guided filtering;

[0070] The filtering module is configured to perform spatiotemporal filtering on the first downsampled image and the reference frame images of the corresponding scale in the reference frame image group based on the motion estimation weights at the corresponding scale, so as to obtain the noise-reduced downsampled image;

[0071] The fusion module is configured to upsample the denoised downsampled image to obtain a denoised upsampled image, and then weight and fuse the denoised upsampled image with the high-frequency information according to the texture estimation weights at the corresponding scale to obtain the final image.

[0072] The second decomposition module downsamples the final image and performs Gaussian pyramid decomposition to obtain a group of reference frame images for noise reduction of the next frame image.

[0073] Optionally, the maximum size of the reference frame image group is the image size after the reference frame image has been downsampled for the first time.

[0074] Optionally, the motion estimation can be performed using inter-frame difference, optical flow, or density-based clustering algorithms.

[0075] Optionally, the texture estimation can be performed using an edge extraction operator.

[0076] Based on the same inventive concept, embodiments of the present invention also propose a readable storage medium on which a computer program is stored, which, when executed, can implement the image noise reduction method described above.

[0077] A readable storage medium can be a tangible device capable of holding and storing instructions for use by an instruction execution device, such as, but not limited to, electrical storage devices, magnetic storage devices, optical storage devices, electromagnetic storage devices, semiconductor storage devices, or any suitable combination thereof. More specific examples of readable storage media (a non-exhaustive list) include: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable compact disc read-only memory (CD-ROM), digital multifunction disc (DVD), memory sticks, floppy disks, mechanical encoding devices, such as punch cards or recessed protrusions storing instructions thereon, and any suitable combination thereof. The computer programs described herein can be downloaded from the readable storage medium to various computing / processing devices, or downloaded via a network, such as the Internet, local area network, wide area network, and / or wireless network, to an external computer or external storage device. Networks can include copper transmission cables, fiber optic transmissions, wireless transmissions, routers, firewalls, switches, gateway computers, and / or edge servers. Each computing / processing device's network adapter card or network interface receives and forwards a computer program from the network for storage on a readable storage medium within the respective computing / processing device. The computer program used to perform the operations of this invention can be execution instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, status setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C++, etc., and conventional procedural programming languages ​​such as "C" or similar languages. The computer program can execute entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, electronic circuits, such as programmable logic circuits, field-programmable gate arrays (FPGAs), or programmable logic arrays (PLAs), are personalized by utilizing state information from a computer program. These electronic circuits can execute computer-readable program instructions, thereby realizing various aspects of the present invention.

[0078] Various aspects of the present invention are described herein with reference to flowchart illustrations and / or block diagrams of methods, systems, and computer program products according to embodiments of the invention. It should 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 a computer program. These computer programs can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to produce a machine such that, when executed by the processor of the computer or other programmable data processing apparatus, they create means for implementing the functions / actions specified in one or more blocks of the flowchart illustrations and / or block diagrams. These computer programs can also be stored in a readable storage medium that causes a computer, programmable data processing apparatus, and / or other device to operate in a particular manner; thus, the readable storage medium storing the computer program comprises an article of manufacture including instructions for implementing aspects of the functions / actions specified in one or more blocks of the flowchart illustrations and / or block diagrams.

[0079] A computer program may also be loaded onto a computer, other programmable data processing apparatus, or other device to cause a series of operational steps to be performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process, thereby causing the computer program executing on the computer, other programmable data processing apparatus, or other device to perform the functions / actions specified in one or more boxes of a flowchart and / or block diagram.

[0080] In summary, the embodiments of the present invention provide an image denoising method, system, and readable storage medium. By performing motion estimation and texture estimation on multi-scale images and correcting the weights at each scale through guided filtering, more accurate motion estimation weights and texture estimation weights can be obtained. These weights are then used for spatiotemporal filtering and fusion of high-frequency information, increasing the image signal-to-noise ratio while preserving more texture details, resulting in a denoising result with better signal-to-noise ratio and richer texture. Furthermore, the reference frame image group used for denoising the next frame is obtained by downsampling the current image and performing Gaussian pyramid decomposition, effectively saving transmission bandwidth.

[0081] The above description is merely a description of preferred embodiments of the present invention and is not intended to limit the scope of the invention in any way. Any changes or modifications made by those skilled in the art based on the above disclosure are within the protection scope of the present invention. Obviously, those skilled in the art can make various modifications and variations to the present invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the present invention and its equivalents, the present invention also intends to include these modifications and variations.

Claims

1. An image denoising method, characterized in that, include: The current frame image is subjected to Gaussian filtering and first downsampling to obtain the first downsampled image. The first downsampled image is then subjected to Gaussian pyramid decomposition to obtain a current frame image group including multiple scale images. The first downsampled image is upsampled to obtain an upsampled image, and the high-frequency information of the current frame image is obtained by subtracting the current frame image from the upsampled image. Motion and texture estimation are performed on the current frame image group and the pre-cached reference frame image group at different scales, and the motion estimation weights and texture estimation weights at each scale are corrected layer by layer from the smallest scale to the largest scale through guided filtering. Based on the motion estimation weights at the corresponding scale, the first downsampled image and the reference frame images at the corresponding scale in the reference frame image group are subjected to spatiotemporal filtering to obtain the noise-reduced downsampled image; The downsampled image after denoising is upsampled to obtain a denoised upsampled image. The upsampled image after denoising is then weighted and fused with the high-frequency information according to the texture estimation weight at the corresponding scale to obtain the final image. The final image is downsampled and decomposed into a Gaussian pyramid to obtain a reference frame image group for noise reduction of the next frame image; wherein, the maximum size of the reference frame image group is the image size of the reference frame image after the first downsampling.

2. The image denoising method according to claim 1, characterized in that, The motion estimation is performed using the inter-frame difference method, optical flow method, or density-based clustering algorithm.

3. The image denoising method according to claim 1, characterized in that, The texture estimation is performed using an edge extraction operator.

4. An image noise reduction system, characterized in that, include: The first decomposition module is configured to perform Gaussian filtering on the current frame image and perform a first downsampling to obtain a first downsampled image, and perform Gaussian pyramid decomposition on the first downsampled image to obtain a current frame image group including multiple scale images. The upsampling module is configured to upsample the first downsampled image to obtain an upsampled image, and to subtract the current frame image from the upsampled image to obtain the high-frequency information of the current frame image; The estimation module is configured to perform motion estimation and texture estimation at different scales on the current frame image group and the pre-cached reference frame image group, and to correct the motion estimation weights and texture estimation weights at each scale layer by layer from the smallest scale to the largest scale through guided filtering; The filtering module is configured to perform spatiotemporal filtering on the first downsampled image and the reference frame images of the corresponding scale in the reference frame image group based on the motion estimation weights at the corresponding scale, so as to obtain the noise-reduced downsampled image; The fusion module is configured to upsample the denoised downsampled image to obtain a denoised upsampled image, and then weight and fuse the denoised upsampled image with the high-frequency information according to the texture estimation weights at the corresponding scale to obtain the final image. The second decomposition module downsamples the final image and performs Gaussian pyramid decomposition to obtain a reference frame image group for noise reduction of the next frame image; wherein, the maximum size of the reference frame image group is the image size of the reference frame image after the first downsampling.

5. The image noise reduction system according to claim 4, characterized in that, The motion estimation is performed using the inter-frame difference method, optical flow method, or density-based clustering algorithm.

6. The image noise reduction system according to claim 4, characterized in that, The texture estimation is performed using an edge extraction operator.

7. A readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it can implement the image noise reduction method according to any one of claims 1-3.

Images