A method, device and computer storage medium for video image noise reduction

Abstract

The present invention provides a video image noise reduction method, apparatus and computer storage medium. The method includes: statistically analyzing the pixel difference between the current frame image data and the corresponding image data of the previous frame after filtering and noise reduction; calculating the gradient information of the image data of the previous frame after filtering and noise reduction; determining a first filtering coefficient for fusing the current frame and the previous frame based on the pixel difference and the gradient information; sequentially performing spatial domain filtering and time domain filtering on the first filtering coefficient to obtain a second filtering coefficient; fusing the current frame image data and the corresponding image data of the previous frame after filtering and noise reduction according to the second filtering coefficient to obtain the image data of the current frame after filtering and noise reduction; storing the second filtering coefficient and the image data of the current frame after filtering and noise reduction for the filtering and noise reduction processing of the next frame of image data. The method is simple to implement and has good video noise reduction effect.

Description

Technical Field

[0001] The present invention belongs to the field of video image processing technology, and in particular, relates to a video image noise reduction method, device, and computer storage medium. Background Art

[0002] Image blurring caused by optical systems, motion, etc. during the formation, transmission, and storage of images, as well as noise from circuit and optical factors, degrades the image quality. However, noise mainly comes from the image acquisition and transmission processes. First, the operating conditions of the image sensor are affected by various factors, such as the environmental conditions during image acquisition and the quality of the sensing components themselves. For example, when using a CCD camera to acquire images, the light intensity and the temperature of the sensor are the main factors generating a large amount of noise in the generated images. Second, the image is mainly contaminated by noise during the transmission process due to interference in the used transmission channel. For instance, an image transmitted through a wireless network may be contaminated due to interference from light or atmospheric factors.

[0003] Generally speaking, noise not only reduces the quality of the acquired image data but also has a negative impact on other modules in the image processing pipeline. Therefore, a filter that can reduce noise while not affecting image details is crucial for the imaging system. Currently, the current video noise reduction processing has poor noise reduction effects under complex conditions and there is an obvious ghosting phenomenon.

[0004] How to achieve an ideal video noise reduction effect under complex conditions is an urgent problem to be solved currently. Summary of the Invention

[0005] In view of this, the present invention provides a video image noise reduction method, device, and computer storage medium to solve the deficiencies of the prior art.

[0006] To achieve the above object, the present invention adopts the following technical solutions:

[0007] An embodiment of the present invention provides a video image noise reduction method, including:

[0008] Statistically calculate the pixel difference between the current frame image data and the corresponding image data of the previous frame after filtering and noise reduction;

[0009] Calculate the gradient information of the image data of the previous frame after filtering and noise reduction;

[0010] Determine the first filtering coefficient for fusing the current frame and the previous frame based on the pixel difference and the gradient information;

[0011] Perform spatial domain filtering and temporal domain filtering on the first filtering coefficient in sequence to obtain a second filtering coefficient;

[0012] Fuse the current frame image data and the corresponding image data after filtering and noise reduction of the previous frame according to the second filtering coefficient to obtain the image data after filtering and noise reduction of the current frame;

[0013] Store the second filtering coefficient and the image data after filtering and noise reduction of the current frame for the filtering and noise reduction processing of the next frame of image data.

[0014] Further, the input image data format is RAW format or YUV format.

[0015] Further, calculate the gradient information of the image data after filtering and noise reduction of the previous frame by performing neighborhood convolution on the image data after filtering and noise reduction of the previous frame using two direction templates, one of the two direction templates detects horizontal edges and the other detects vertical edges.

[0016] Further, determining the first filtering coefficient for fusing the current frame and the previous frame from the pixel difference and the gradient information specifically includes:

[0017] Calculate the noise intensity characterization value and the motion intensity characterization value according to the pixel difference;

[0018] Determine the pixel difference gain and the gradient information gain according to the values of the noise intensity characterization value and the motion intensity characterization value using a preset mapping look-up table;

[0019] Determine the sum of the result of multiplying the pixel difference gain by the pixel difference and the result of multiplying the gradient information gain by the gradient information as the first filtering coefficient for fusing the current frame and the previous frame.

[0020] Further, calculating the noise intensity characterization factor and the motion intensity characterization factor according to the pixel difference specifically includes:

[0021] Perform a histogram statistics on the pixel absolute difference;

[0022] Calculate the sum of the histogram in the interval [0, k]. When the proportion of the sum of the interval [0, k] in the total sum of the entire histogram reaches a preset percentage, record the current k value, and this k value is counted as the noise intensity characterization value;

[0023] Divide the image into multiple blocks of the same size, and respectively count the number of pixel absolute differences greater than a preset threshold value in each image block, and this number is counted as the motion intensity characterization value of each image block.

[0024] Further, the preset mapping look-up table divides the noise intensity characterization value into X + 1 levels through X different noise intensity threshold values; the preset mapping look-up table divides the motion intensity characterization value into Y + 1 levels through Y different motion intensity threshold values, thereby mapping out (X + 1)(Y + 1) groups of corresponding pixel difference gains and gradient information gains.

[0025] Further, the spatial domain filtering of the filtering coefficients specifically includes: performing low-pass filtering on the filtering coefficients to filter out the interference of noise; and then performing maximum filtering on the result of the low-pass filtering to retain the details of the image.

[0026] The time domain filtering of the filtering coefficients specifically includes: when the current frame coefficient calculated in the current filtering process is smaller than the filtering coefficient calculated and stored in the previous filtering process; the filtering coefficient of the current frame in the current filtering process is weighted using the following expression: alpha(i,x,y)' = K * alpha(i,x,y) + (1 - K) * alpha(i - 1,x,y); where K is a preset coefficient, alpha(i,x,y) represents the filtering coefficient at the current frame (x,y) in the current filtering process, alpha(i - 1,x,y) represents the filtering coefficient at the current frame (x,y) calculated and stored in the previous filtering process, and alpha(i,x,y)' represents the filtered coefficient of the current frame after weighting.

[0027] Further, the current frame image data and the corresponding image data after filtering and noise reduction of the previous frame are fused according to the second filtering coefficient to obtain the image data after filtering and noise reduction of the current frame using the following expression:

[0028] If(i) = alpha(i).*I(i) + (1 - alpha(i)).*If(i - 1), where alpha(i) is the second filtering coefficient, I(i) is the current frame image data, If(i - 1) is the corresponding image data after filtering and noise reduction of the previous frame, and If(i) is the corresponding image data after filtering and noise reduction of the current frame.

[0029] An embodiment of the present invention provides a video image noise reduction device, including:

[0030] A difference statistics module, configured to statistically calculate the pixel difference between the current frame image data and the corresponding image data after filtering and noise reduction of the previous frame;

[0031] A gradient calculation module, configured to calculate the gradient information of the image data after filtering and noise reduction of the previous frame;

[0032] A coefficient determination module, configured to determine the first filtering coefficient for fusing the current frame and the previous frame based on the pixel difference and the gradient information;

[0033] A coefficient correction module, configured to perform spatial domain filtering and time domain filtering on the first filtering coefficient in sequence to obtain the second filtering coefficient;

[0034] A fusion filtering module, configured to fuse the current frame image data and the corresponding image data after filtering and noise reduction of the previous frame according to the second filtering coefficient to obtain the image data after filtering and noise reduction of the current frame;

[0035] A data storage module, configured to store the second filtering coefficient and the image data after filtering and noise reduction of the current frame for the filtering and noise reduction processing of the next frame of image data.

[0036] An embodiment of the present invention also provides a computer storage medium, on which a computer program is stored. When the computer program is executed by a processor, the above-mentioned video image noise reduction method is implemented.

[0037] The video image noise reduction method provided by the present invention performs denoising based on gradient information and difference statistics, is easy to be implemented by hardware, has good noise reduction effect, less ghosting, and can greatly improve the quality of video data. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0039] Figure 1 It is a schematic flowchart of a video image noise reduction method provided by an embodiment of the present invention;

[0040] Figure 2 It is a schematic flowchart of determining a filtering coefficient from pixel differences and gradient information;

[0041] Figure 3 It is a schematic structural diagram of a video image noise reduction device provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0042] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0043] Figure 1 Shown is a schematic flowchart of a video image noise reduction method provided by an embodiment of the present invention.

[0044] S1: Input the current frame data.

[0045] The input image data format is applicable to RAW format or YUV format.

[0046] The noise reduction method provided by the present invention can be deployed before or after the demosaicing module on the ISP pipeline according to requirements. If it is deployed before the demosaicing module, the input data is in RAW data format. If it is deployed after the demosaicing module, the input data is in YUV data format. Denote the current frame image data as I(i).

[0047] S2: Statistically calculate the pixel differences between the current frame image data and the corresponding image data after filtering and noise reduction of the previous frame.

[0048] Statistically calculate the pixel differences between two adjacent frames. The statistical calculation of pixel differences is to calculate the absolute difference of each corresponding pixel between the current frame data and the data after filtering of the previous frame. The differences between two pixels at the same coordinates in two adjacent frames are mainly caused by the movement and noise in the picture. Denote the image data after filtering and noise reduction of the previous frame as If(i - 1), then the pixel absolute difference is denoted as Diff = abs(I(i) - If(i - 1)).

[0049] S3: Calculate the gradient information of the image data after filtering and noise reduction of the previous frame.

[0050] The calculation of the gradient information of the image data after filtering and noise reduction of the previous frame is achieved by performing neighborhood convolution on the image data after filtering and noise reduction of the previous frame using two direction templates. One of the two direction templates detects horizontal edges, and the other detects vertical edges.

[0051] Calculate the gradient information of the data after filtering of the previous frame. The gradient information consists of a horizontal gradient and a vertical gradient. The horizontal gradient operator is Gx = [-1 0 1; -1 0 1; -1 0 1], and the vertical gradient operator is Gy = [-1 -1 -1; 0 0 0; 1 1 1]. The gradient Grad = abs(If(i - 1) * Gx) + abs(If(i - 1) * Gy), where the asterisk "*" represents convolution.

[0052] S4: Determine the filtering coefficient for fusing the current frame and the previous frame based on the pixel differences and the gradient information.

[0053] The filtering coefficient is jointly determined by the pixel difference information and the gradient information. The specific steps are as Figure 2 shown

[0054] S401: Calculate the noise intensity characterization value and the motion intensity characterization value according to the pixel differences.

[0055] The calculation of the noise intensity characterization value and the motion intensity characterization value based on the pixel difference can be specifically carried out in the following manner: First, perform a histogram statistics on the pixel absolute difference Diff; for example: a 32-order histogram statistics can be performed on the pixel absolute difference Diff. Then calculate the sum of the histogram in the interval [0, k]. When the proportion of the sum of this interval in the total sum of the entire histogram reaches the preset percentage, record the current k value, and the k value is counted as the noise intensity characterization value; the preset percentage is an engineering experience value. Usually, the value range is 60%-75%. In this embodiment, the preset percentage is preferably 68%. The larger the k value, the greater the noise. Finally, divide the image into multiple blocks of the same size, and respectively count the number Num_mov of pixel absolute differences Diff greater than the preset threshold value th in each block, where the preset threshold value th is determined by the k value. The larger the k value, the larger the th value. Num_mov is counted as the motion intensity characterization value of each block.

[0056] S402: Determine the pixel difference gain and the gradient information gain according to the values of the noise intensity and the motion intensity characterization value by using a preset mapping look-up table.

[0057] The preset mapping look-up table divides the noise intensity into X + 1 levels through X different noise intensity threshold values; the preset mapping look-up table divides the motion levels into Y + 1 levels through Y different motion intensity threshold values, so as to map (X + 1)(Y + 1) groups of corresponding pixel difference gains and gradient information gains. The preset mapping look-up table is obtained through multiple experimental tests.

[0058] For example: when X = 2 and Y = 2, we can estimate and classify the noise intensity of the video into the following three categories according to the noise intensity (the k value obtained in step S401): environmentally friendly, appropriate lighting, and basically no noise in the video; general lighting environment, and there is a certain amount of noise in the video; low lighting, and there is a large amount of noise in the video. We can estimate and classify the motion intensity into the following three categories according to the motion intensity (the Num_mov value obtained in step S401): almost no motion; weak motion; strong motion. When there is basically no noise in the video, regardless of the current scene motion intensity, make the difference gain Gd and the gradient information gain Gg the smallest, such as Gd = 0, Gg = 0. When there is a certain amount of noise in the video but almost no motion, make the difference gain larger and the gradient gain smaller, such as Gd = 4, Gg = 1. When there is a certain amount of noise in the video and for blocks with weak motion, make the difference gain larger and the gradient gain larger, such as Gd = 4, Gg = 4. For blocks with strong motion, make the difference gain very large and the gradient gain very large, such as Gd = 8, Gg = 7. When there is a large amount of noise in the video and almost no motion, make the difference gain smaller and the gradient gain very small, such as Gd = 1, Gg = 0. When there is a large amount of noise in the video and for blocks with weak motion, make the difference gain smaller and the gradient gain smaller, such as Gd = 2, Gg = 0. For blocks with strong motion, make the difference gain larger and the gradient gain smaller, such as Gd = 3, Gg = 1. Specifically, according to the actual test situation, the situation can be divided into more categories, and different gain values can be set through category judgment, so as to adaptively perform noise reduction processing on different scenes.

[0059] S403: Determine the sum of the result of multiplying the pixel difference gain by the pixel difference and the result of multiplying the gradient information gain by the gradient information as the filtering coefficient for fusing the current frame and the previous frame.

[0060] The calculation of the filtering coefficient uses the following expression:

[0061] alpha(i) = Gd.*Diff + Gg.*Grad, where ".*" represents dot product; alpha(i) represents the filtering coefficient; Gd is the difference gain, Gg is the gradient information gain; Diff is the pixel absolute difference, and Grad is the gradient information; the filtering coefficient obtained by dot product addition is the initial filtering coefficient. Subsequent spatial and temporal filtering processing is also required.

[0062] S5: Perform spatial domain filtering on the filtering coefficient.

[0063] First, perform low-pass filtering on the filtering coefficient alpha(i) to filter out noise interference, and then perform maximum filtering on the result of the low-pass filtering to retain the details of the image. Through low-pass linear filtering and maximum non-linear filtering, while blurring and reducing noise, the sharpness of edges and details in the image can be ensured without loss. In this embodiment, the low-pass filtering can be a 3*3 Gaussian low-pass filter or a 5*5 Gaussian filter, and the maximum filtering can be a 3*7, 5*5, or 5*7 maximum filter. Similarly, it is adaptively selected according to the noise estimation intensity and motion estimation intensity.

[0064] S6: Perform time-domain filtering on the filtering coefficient.

[0065] After the filtering coefficient is subjected to spatial-domain filtering, it is then subjected to time-domain filtering with the filtering coefficient alpha(i - 1) calculated and stored in the previous frame. Specifically: alpha(i,x,y) represents the filtering coefficient at the current frame (x,y) in the current filtering process, and alpha(i - 1,x,y) represents the filtering coefficient at the current frame (x,y) in the previous filtering process. If alpha(i,x,y) < alpha(i - 1,x,y), that is, when the coefficient of the current frame calculated in the current filtering process is smaller than the coefficient of the current frame calculated in the previous filtering process, the filtering coefficient at the current frame (x,y) in the current filtering process is weighted: alpha(i,x,y)' = K * alpha(i,x,y) + (1 - K) * alpha(i - 1,x,y); the value of K can be selected from

[01] , and is specifically selected according to experience, with different values configured for different scenarios. alpha(i,x,y)' represents the filtered coefficient of the current frame after weighting. No other processing is performed in other cases.

[0066] S7: According to the filtering coefficient, fuse the current frame image data and the corresponding image data after filtering and noise reduction in the previous frame to obtain the image data after filtering and noise reduction of the current frame.

[0067] Perform time-domain filtering on two adjacent frames according to the processed coefficient to obtain the filtering result of the current frame If(i) = alpha(i).*I(i) + (1 - alpha(i)).*If(i - 1), where alpha(i) is the filtering coefficient, I(i) is the current frame image data, If(i - 1) is the image data after filtering and noise reduction of the previous frame, and If(i) is the image data after filtering and noise reduction of the current frame.

[0068] S8: Store the filtering coefficient and the image data after filtering and noise reduction of the current frame.

[0069] Store the filtering coefficient and the image data after filtering and noise reduction of the current frame for the filtering and noise reduction processing of the next frame of image data. Store the filtering coefficient alpha(i) and the filtering result If(i) for the next filtering.

[0070] S9: Output the filtering result.

[0071] Output the filtering result If(i) to the next processing module.

[0072] Figure 3 The figure shows a schematic structural diagram of a video image noise reduction device provided by an embodiment of the present invention. The video image noise reduction device includes: a difference statistics module, a gradient calculation module, a coefficient determination module, a coefficient correction module, a fusion filtering module, and a data storage module.

[0073] The difference statistics module is used to statistically calculate the pixel difference between the current frame image data and the corresponding image data of the previous frame after filtering and noise reduction; the gradient calculation module is used to calculate the gradient information of the image data of the previous frame after filtering and noise reduction; the coefficient determination module is used to determine the first filtering coefficient for fusing the current frame and the previous frame based on the pixel difference and the gradient information; the coefficient correction module is used to perform spatial domain filtering and temporal domain filtering on the first filtering coefficient in sequence to obtain the second filtering coefficient; the fusion filtering module is used to fuse the current frame image data and the corresponding image data of the previous frame after filtering and noise reduction according to the second filtering coefficient to obtain the image data of the current frame after filtering and noise reduction; the data storage module is used to store the second filtering coefficient and the image data of the current frame after filtering and noise reduction for the filtering and noise reduction processing of the next frame of image data.

[0074] It should be noted that: when the above-mentioned video noise reduction device provided in the embodiment performs noise reduction, only the above-mentioned division of each program module is used for illustration. In practical applications, the above-mentioned processing can be allocated to different program modules according to needs, that is, the internal structure of the device is divided into different program modules to complete all or part of the above-described processing. In addition, the above-mentioned video noise reduction device provided in the embodiment and the embodiment of the video image noise reduction method belong to the same concept. For the specific implementation process, please refer to the method embodiment. The beneficial effects are the same as those of the method embodiment and will not be elaborated here.

[0075] An embodiment of the present invention further provides a computer storage medium, which is a computer-readable storage medium, storing a computer program thereon. The computer program can be executed by a processor of the video noise reduction device to complete the steps of the foregoing video image noise reduction method. The computer-readable storage medium can be a ferromagnetic random access memory (FRAM), a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), a flash memory, a magnetic surface memory, an optical disc, or a compact disc read-only memory (CD-ROM), etc.

[0076] In several embodiments provided by the present invention, it should be understood that the disclosed methods and intelligent devices can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined, or integrated into another system, or some features can be ignored, or not executed. In addition, the coupling, direct coupling, or communication connection between the displayed or discussed components can be through some interfaces. The indirect coupling or communication connection of the devices or units can be electrical, mechanical, or other forms.

[0077] The above is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, which should be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

1. A video image noise reduction method, characterized in that, Including: Statistically calculate the pixel difference between the current frame image data and the corresponding image data after filtering and noise reduction of the previous frame; Calculate the gradient information of the image data after filtering and noise reduction of the previous frame; Determine the first filtering coefficient for fusing the current frame and the previous frame based on the pixel difference and the gradient information; Perform spatial domain filtering and temporal domain filtering on the first filtering coefficient in sequence to obtain the second filtering coefficient; Fuse the current frame image data and the corresponding image data after filtering and noise reduction of the previous frame according to the second filtering coefficient to obtain the image data after filtering and noise reduction of the current frame; Store the second filtering coefficient and the image data after filtering and noise reduction of the current frame for the filtering and noise reduction processing of the next frame of image data.

2. The video image noise reduction method according to claim 1, wherein, The input image data format is RAW format or YUV format.

3. The video image noise reduction method according to claim 1, characterized in that, Calculating the gradient information of the image data after filtering and noise reduction of the previous frame is achieved by performing neighborhood convolution on the image data after filtering and noise reduction of the previous frame using two directional templates, where one of the two directional templates detects horizontal edges and the other detects vertical edges.

4. The video image noise reduction method according to claim 1, characterized in that, Determining the first filtering coefficient for fusing the current frame and the previous frame based on the pixel difference and the gradient information specifically includes: Calculate the noise intensity characterization value and the motion intensity characterization value according to the pixel difference; Determine the pixel difference gain and the gradient information gain using a preset mapping look-up table based on the values of the noise intensity and the motion intensity characterization value; Determine the sum of the result of multiplying the pixel difference gain by the pixel difference and the result of multiplying the gradient information gain by the gradient information as the first filtering coefficient for fusing the current frame and the previous frame.

5. The video image noise reduction method according to claim 4, wherein, Calculating the noise intensity characterization factor and the motion intensity characterization factor according to the pixel difference specifically includes: Perform histogram statistics on the absolute pixel difference; Calculate the sum of the histogram in the interval [0, k]. When the proportion of the sum of the interval [0, k] in the total sum of the entire histogram reaches a preset percentage, record the current k value, and this k value is counted as the noise intensity characterization value; Divide the image into multiple blocks of the same size, and respectively statistically calculate the number of pixels with absolute difference greater than a preset threshold in each image block, and this number is counted as the motion intensity characterization value of each image block.

6. The video image noise reduction method according to claim 4, wherein The preset mapping look-up table divides the noise intensity characterization value into X + 1 levels through X different noise intensity threshold values; the preset mapping look-up table divides the motion intensity characterization value into Y + 1 levels through Y different motion intensity threshold values, thereby mapping out (X + 1)(Y + 1) groups of corresponding pixel difference gains and gradient information gains.

7. The video image noise reduction method according to claim 1, wherein Performing spatial domain filtering on the filtering coefficient specifically includes: performing low-pass filtering on the filtering coefficient to filter out the interference of noise; then performing maximum value filtering on the result of the low-pass filtering to retain the details of the image; The time-domain filtering of the filtering coefficient specifically includes: when the current frame coefficient calculated in the current filtering process is smaller than the filtering coefficient calculated and stored in the previous filtering process; the filtering coefficient of the current frame in the current filtering process is weighted using the following expression: alpha(i,x,y)' = K * alpha(i,x,y) + (1 - K) * alpha(i - 1,x,y); where K is a preset coefficient, alpha(i,x,y) represents the filtering coefficient at the current frame (x,y) in the current filtering process, alpha(i - 1,x,y) represents the filtering coefficient at the current frame (x,y) calculated and stored in the previous filtering process, and alpha(i,x,y)' represents the filtered coefficient of the current frame after weighting.

8. The video image noise reduction method according to claim 1, wherein The current frame image data and the corresponding image data after filtering and noise reduction of the previous frame are fused according to the second filtering coefficient to obtain the image data after filtering and noise reduction of the current frame using the following expression: If(i) = alpha(i).*I(i) + (1 - alpha(i)).*If(i - 1), where alpha(i) is the second filtering coefficient, I(i) is the current frame image data, If(i - 1) is the corresponding image data after filtering and noise reduction of the previous frame, and If(i) is the corresponding image data after filtering and noise reduction of the current frame.

9. A video image noise reduction device, characterized in that, It includes: A difference statistics module for statistically calculating the pixel difference between the current frame image data and the corresponding image data after filtering and noise reduction of the previous frame; A gradient calculation module for calculating the gradient information of the image data after filtering and noise reduction of the previous frame; A coefficient determination module for determining the first filtering coefficient for fusing the current frame and the previous frame based on the pixel difference and the gradient information; A coefficient correction module for sequentially performing spatial-domain filtering and time-domain filtering on the first filtering coefficient to obtain the second filtering coefficient; A fusion filtering module for fusing the current frame image data and the corresponding image data after filtering and noise reduction of the previous frame according to the second filtering coefficient to obtain the image data after filtering and noise reduction of the current frame; A data storage module for storing the second filtering coefficient and the image data after filtering and noise reduction of the current frame for the filtering and noise reduction processing of the next frame of image data.

10. A computer storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the video image noise reduction method according to any one of claims 1 to 8.

Images