CN103702016A - Video denoising method and device - Google Patents
Video denoising method and device Download PDFInfo
- Publication number
- CN103702016A CN103702016A CN201310712320.5A CN201310712320A CN103702016A CN 103702016 A CN103702016 A CN 103702016A CN 201310712320 A CN201310712320 A CN 201310712320A CN 103702016 A CN103702016 A CN 103702016A
- Authority
- CN
- China
- Prior art keywords
- pixel
- component
- result
- weighted statistical
- noise reduction
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Landscapes
- Picture Signal Circuits (AREA)
Abstract
The invention provides a video denoising method and a video denoising device. The method comprises the following steps of: converting a video streaming signal into YUV three components, and acquiring the Y component of a pixel in each frame of image in a video streaming signal; carrying out weighting statistical calculation on the Y component of each pixel; judging whether noise exists on the pixel according to the result of the weighting statistical calculation and the Y component of the relevant pixel; if yes, denoising the relevant pixel according to the result of the weighting statistical calculation. According to the video denoising method and the video denoising device, a little of calculated amount and lesser hardware resources can be only needed in the denoising process, so as to obtain a good denoising effect.
Description
Technical field
The present invention relates to technical field of image processing, be specifically related to a kind of vedio noise reduction method and a kind of vedio noise reduction device.
Background technology
The object of vedio noise reduction is exactly in order to reduce video noise.Video noise comes from many-side.Video noise is by coming from device external disturbance, as electromagnetic wave and external noise inner through power supply string motion device and that cause.Also have the interference that comes from device inside, as the thermal noise of video camera, electrical machinery moves and the internal noises such as jittering noise of generation.If can not effectively remove these noises, by having a strong impact on the subjective quality of video image, reduce video compression efficiency simultaneously.Noise brings more high fdrequency component can to the residual error of Intra piece and Inter piece, thereby needs more bit to retain these unwanted information.In addition, the existence of noise makes the match block of searching for current block in reference frame become more difficult.In other words, if video compression is arrived to low code check, noise will cause its Subjective and objective qualities obviously to decline.Therefore necessary video is carried out to noise reduction process.
Existing vedio noise reduction method mainly contains pixel domain filtering processing and time-domain filtering is processed two kinds.Wherein, pixel domain filter generally carries out in the operation window of the specific size centered by current pixel, as harmonic wave mean filter, weighted arithmetic mean filtering, the filtering of α-trimmed mean and medium filtering etc.These class methods are very large to dissimilar noiseproof feature performance difference.For example, mean filter is one of best filter of removing Gaussian noise, edge and details that but can fuzzy objective.Median filter adopts the intermediate value of surrounding pixel to replace current pixel value, can effectively remove impulsive noise, avoids edge blurry simultaneously, but is not fine to the removal effect of individual Gaussian noise.Pixel domain filter effect and intensity have very large relation, and intensity too conference causes edge blurry and loss in detail, too little not thorough to the removal of noise.
In addition, time-domain filtering is mainly by motion compensation technique, at the motion of time domain tracking object filtering noise.Existence due to noise, by minimizing the blocks and optimal matching blocks that the method for searching motion of residual absolute value sum finds, it not often the position that actual object is corresponding, if directly carried out based on this, filtering will cause and obvious edge blurry and loss in detail, especially work as the larger situation of power ratio of noise.
The amount of calculation needing due to existing vedio noise reduction method is larger, thereby needs to consume larger hardware resource, causes noise reduction poor.
Summary of the invention
Based on this, the invention provides a kind of vedio noise reduction method and device, can effectively improve vedio noise reduction effect.
For achieving the above object, the present invention adopts following technical scheme:
A vedio noise reduction method, comprises the following steps:
Video stream signal is changed into YUV three-component, and obtain the Y component of pixel in interior each two field picture of video stream signal;
Y component to each pixel is weighted statistical computation;
The result of calculating according to described weighted statistical and the Y component of respective pixel judge whether this pixel exists noise;
If so, the result of calculating according to described weighted statistical is carried out noise reduction process to respective pixel.
A vedio noise reduction device, comprising:
Component acquisition module, for video stream signal being changed into YUV three-component, and obtains in video stream signal the Y component of pixel in each two field picture;
Weighted statistical computing module, is weighted statistical computation to the Y component of each pixel;
Judge module, for judging according to the result of described weighted statistical calculating and the Y component of respective pixel whether this pixel exists noise;
Noise processed module, in the situation that judgment result is that of described judge module is that the result of calculating according to described weighted statistical is carried out noise reduction process to respective pixel.
By above scheme, can be found out, a kind of vedio noise reduction method of the present invention and device, by obtaining the three-component Y component of pixel YUV in each two field picture, then Y component is weighted to statistical computation, the result of calculating according to weighted statistical is again carried out noise reduction process to respective pixel, thereby reaches the effect of vedio noise reduction.Because a kind of vedio noise reduction method of the present invention and device are by each two field picture pixel Y component is carried out to statistical weight, it has utilized the incidence relation of each pixel between video sequence to do a weighted statistical and has calculated, so only need amount of calculation seldom and less hardware resource in processing the process of noise, improve vedio noise reduction effect, and can effectively prevent the edge of fuzzy objective and the situation of loss in detail.
Accompanying drawing explanation
Fig. 1 is a kind of vedio noise reduction method flow schematic diagram in the embodiment of the present invention;
Fig. 2 is a kind of vedio noise reduction apparatus structure schematic diagram in the embodiment of the present invention.
Embodiment
In order to make object of the present invention, technical scheme and advantage clearer, below in conjunction with drawings and Examples, the present invention is further elaborated.Should be appreciated that specific embodiment described herein, only in order to explain the present invention, is not intended to limit the present invention.
Shown in Figure 1, a kind of vedio noise reduction method, is characterized in that, comprises the following steps:
Step S101, changes into YUV three-component by video stream signal, and obtains the Y component of pixel in interior each two field picture of video stream signal; YUV is the kind of compiling true-color color space.Wherein " Y " represents the lightness (Luminance, Luma) of pixel, and " U " and " V " represents respectively colourity, the concentration (Chrominance, Chroma) of pixel.It should be noted that, above-mentionedly video stream signal is changed into YUV three-component can adopt prior art, in the present invention, it will not go into details.
Step S102, is weighted statistical computation to the Y component of each pixel.Add up the pixel situation of change of continuous multiple frames video image, from more away from frame, it is just lower to the weighing factor of each pixel of present frame.It should be noted that, in the present invention, only Y component is weighted to statistical computation, and U component and V component remain unchanged.
Step S103, the result of calculating according to described weighted statistical and the Y component of respective pixel judge whether this pixel exists noise; If so, there is noise in explanation, enters step S104; Otherwise illustrate and do not have noise, without carrying out noise reduction process.
Step S104, if there is noise, the result of calculating according to described weighted statistical is carried out noise reduction process to respective pixel.
As a good embodiment, the process that the described Y component to each pixel is weighted statistical computation specifically can comprise as follows:
Can adopt following formula to be weighted statistical computation to the Y component of each pixel:
M
n(i,j)=M
n-1(i,j)*α+β*Y
n(i,j)；
N
n(i,j)=N
n-1(i,j)*α+β*Y
n(i,j)
2；
Wherein, M
n(i, j) and N
n(i, j) represents respectively the result that the weighted statistical of pixel in n two field picture calculates; N is greater than 0 natural number; I, the ranks coordinate that j is pixel; M
0(i, j)=0; N
0(i, j)=0; Y
n(i, j) represents the Y component of n two field picture current pixel; α ∈ [0,1]; β ∈ [0,100].
As a good embodiment, the result of calculating according to described weighted statistical and the Y component of respective pixel judge whether this pixel exists the process of noise specifically can comprise as follows:
The result of calculating according to described weighted statistical and the Y component of respective pixel calculate, and computing formula can be as follows:
L
1=N
n(i,j)-M
n(i,j)
2/255；
L
2=(Y
n(i,j)*255-M
n(i,j))
2/255；
According to result of calculation L
1with L
2whether the pixel that judges corresponding frame there is noise.Can be according to described result of calculation L
1with L
2whether be less than the pixel that predetermined value judges corresponding frame and whether have noise.If so, for example, work as L
1<threshold, and L
2<threshold; Wherein threshold ∈ [12750,25500], illustrates that this frame pixel exists noise.Otherwise this frame pixel does not exist noise, the Y component of pixel is constant.
As a good embodiment, while judging according to result of calculation L1 and L2 whether the pixel of corresponding frame exists judgment result is that of noise to be if above-mentioned, the process that the result of calculating according to described weighted statistical is carried out noise reduction process to respective pixel can comprise as follows:
Obtain the result M that the weighted statistical of the described pixel that has a noise calculates
n(i, j);
By formula Y
n(i, j)=M
nthe result M that (i, j) calculates described weighted statistical
nto the Y component of respective pixel, there is the Y component of the pixel of noise in (i, j) assignment.
Corresponding with a kind of vedio noise reduction method in above-described embodiment one, the embodiment of the present invention also provides a kind of vedio noise reduction device, as shown in Figure 2, comprising:
Weighted statistical computing module 102, is weighted statistical computation to the Y component of each pixel;
Noise processed module 104, in the situation that judgment result is that of described judge module is that the result of calculating according to described weighted statistical is carried out noise reduction process to respective pixel.
As a good embodiment, described weighted statistical computing module specifically can comprise:
The first statistics computing module, for according to formula M
n(i, j)=M
n-1 (i, j) * alpha+beta * Y
n(i, j) calculates the first statistical weight result;
The second statistics computing module, for according to formula N
n(i, j)=N
n-1 (i, j) * alpha+beta * Y
n(i, j)
2calculate the second statistical weight result;
Wherein, M
n(i, j) and N
n(i, j) represents respectively the result that the weighted statistical of pixel in n two field picture calculates; N is greater than 0 natural number; I, the ranks coordinate that j is pixel; M
0(i, j)=0; N
0(i, j)=0; Y
n(i, j) represents the Y component of n two field picture current pixel; α ∈ [0,1]; β ∈ [0,100].
As a good embodiment, described judge module specifically can comprise:
Computing module, for calculating according to the result of described weighted statistical calculating and the Y component of respective pixel, computing formula is as follows:
L
1=N
n(i,j)-M
n(i,j)
2/255；
L
2=(Y
n(i,j)*255-M
n(i,j))
2/255；
Sub-judge module, for judging according to result of calculation L1 and L2 whether the pixel of corresponding frame exists noise.
As a good embodiment, described noise processed module specifically can comprise:
Weighted statistical result of calculation acquisition module, for obtaining the result M of the pixel weighted statistical calculating that has noise
n(i, j);
Assignment module, for the result M that described weighted statistical is calculated
n(i, j) assignment is to the Y component of respective pixel.
Other technical characterictic of above-mentioned a kind of vedio noise reduction device is identical with a kind of vedio noise reduction method of the present invention, and it will not go into details herein.
By above scheme, can find out, a kind of vedio noise reduction method of the present invention and device, by obtaining the three-component Y component of pixel YUV in each two field picture, then Y component is weighted to statistical computation, the result of calculating according to weighted statistical is again carried out noise reduction process to respective pixel, thereby reaches the effect of vedio noise reduction.Because a kind of vedio noise reduction method of the present invention and device are by each two field picture pixel Y component is carried out to statistical weight, it has utilized the incidence relation of each pixel between video sequence to do a weighted statistical and has calculated, so only need amount of calculation seldom and less hardware resource in processing the process of noise, improve vedio noise reduction effect, and can effectively prevent the edge of fuzzy objective and the situation of loss in detail.
Due to amount of calculation in process processing the process of noise of a kind of vedio noise reduction method of the present invention and device and the hardware resource that needs all seldom, for example, so can be applicable to the very limited system of computing capability, in the video conferencing system that hardware resource is limited and requirement of real-time is very high.
It should be noted that, unless context separately has the description of specific distinct, the element in the present invention and assembly, the form that quantity both can be single exists, and form that also can be a plurality of exists, and the present invention does not limit this.In addition, although the step in the present invention is arranged with label, but and be not used in the precedence that limits step, unless expressly stated the order of step or the execution of certain step need other steps as basis, otherwise the relative order of step is adjustable.
The above embodiment has only expressed several execution mode of the present invention, and it describes comparatively concrete and detailed, but can not therefore be interpreted as the restriction to the scope of the claims of the present invention.It should be pointed out that for the person of ordinary skill of the art, without departing from the inventive concept of the premise, can also make some distortion and improvement, these all belong to protection scope of the present invention.Therefore, the protection range of patent of the present invention should be as the criterion with claims.
Claims (8)
1. a vedio noise reduction method, is characterized in that, comprises the following steps:
Video stream signal is changed into YUV three-component, and obtain the Y component of pixel in interior each two field picture of video stream signal;
Y component to each pixel is weighted statistical computation;
The result of calculating according to described weighted statistical and the Y component of respective pixel judge whether this pixel exists noise;
If so, the result of calculating according to described weighted statistical is carried out noise reduction process to respective pixel.
2. vedio noise reduction method according to claim 1, is characterized in that, the process that the described Y component to each pixel is weighted statistical computation comprises:
Adopt following formula to be weighted statistical computation to the Y component of each pixel:
M
n(i,j)=M
n-1(i,j)*α+β*Y
n(i,j)；
N
n(i,j)=N
n-1(i,j)*α+β*Y
n(i,j)
2；
Wherein, M
n(i, j) and N
n(i, j) represents respectively the result that the weighted statistical of pixel in n two field picture calculates; N is greater than 0 natural number; I, the ranks coordinate that j is pixel; M
0(i, j)=0; N
0(i, j)=0; Y
n(i, j) represents the Y component of n two field picture current pixel; α ∈ [0,1]; β ∈ [0,100].
3. vedio noise reduction method according to claim 2, is characterized in that, the result of calculating according to described weighted statistical and the Y component of respective pixel judge whether this pixel exists the process of noise to comprise:
The result of calculating according to described weighted statistical and the Y component of respective pixel calculate, and computing formula is as follows:
L1=N
n(i,j)-M
n(i,j)
2/255；
L
2=(Y
n(i,j)*255-M
n(i,j))
2/255；
According to result of calculation L
1with L
2whether the pixel that judges corresponding frame there is noise.
4. according to the vedio noise reduction method described in claim 2 or 3, it is characterized in that, the process that the result of calculating according to described weighted statistical is carried out noise reduction process to respective pixel comprises:
Obtain the result M of the pixel weighted statistical calculating that has noise
n(i, j);
The result M that described weighted statistical is calculated
n(i, j) assignment is to the Y component of respective pixel.
5. a vedio noise reduction device, is characterized in that, comprising:
Component acquisition module, for video stream signal being changed into YUV three-component, and obtains in video stream signal the Y component of pixel in each two field picture;
Weighted statistical computing module, is weighted statistical computation to the Y component of each pixel;
Judge module, for judging according to the result of described weighted statistical calculating and the Y component of respective pixel whether this pixel exists noise;
Noise processed module, in the situation that judgment result is that of described judge module is that the result of calculating according to described weighted statistical is carried out noise reduction process to respective pixel.
6. vedio noise reduction device according to claim 5, is characterized in that, described weighted statistical computing module comprises:
The first statistics computing module, for according to formula M
n(i, j)=M
n-1 (i, j) * alpha+beta * Y
n(i, j) calculates the first statistical weight result;
The second statistics computing module, for according to formula N
n(i, j)=N
n-1 (i, j) * alpha+beta * Y
n(i, j)
2calculate the second statistical weight result;
Wherein, M
n(i, j) and N
n(i, j) represents respectively the result that the weighted statistical of pixel in n two field picture calculates; N is greater than 0 natural number; I, the ranks coordinate that j is pixel; M
0(i, j)=0; N
0(i, j)=0; Y
n(i, j) represents the Y component of n two field picture current pixel; α ∈ [0,1]; β ∈ [0,100].
7. vedio noise reduction device according to claim 6, is characterized in that, described judge module comprises:
Computing module, for calculating according to the result of described weighted statistical calculating and the Y component of respective pixel, computing formula is as follows:
L
1=N
n(i,j)-M
n(i,j)
2/255；
L
2=(Y
n(i,j)*255-M
n(i,j))
2/255；
Sub-judge module, for judging according to result of calculation L1 and L2 whether the pixel of corresponding frame exists noise.
8. according to the vedio noise reduction device described in claim 6 or 7, it is characterized in that, described noise processed module comprises:
Weighted statistical result of calculation acquisition module, for obtaining the result Mn (i, j) of the pixel weighted statistical calculating that has noise;
Assignment module, for the result M that described weighted statistical is calculated
n(i, j) assignment is to the Y component of respective pixel.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310712320.5A CN103702016B (en) | 2013-12-20 | 2013-12-20 | Vedio noise reduction method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310712320.5A CN103702016B (en) | 2013-12-20 | 2013-12-20 | Vedio noise reduction method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103702016A true CN103702016A (en) | 2014-04-02 |
| CN103702016B CN103702016B (en) | 2017-06-09 |
Family
ID=50363404
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310712320.5A Expired - Fee Related CN103702016B (en) | 2013-12-20 | 2013-12-20 | Vedio noise reduction method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103702016B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112805990A (en) * | 2018-11-15 | 2021-05-14 | 深圳市欢太科技有限公司 | Video processing method and device, electronic equipment and computer readable storage medium |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101001317A (en) * | 2007-01-22 | 2007-07-18 | 华为技术有限公司 | Method and device of vedio noise reduction |
| CN101448077A (en) * | 2008-12-26 | 2009-06-03 | 四川虹微技术有限公司 | Self-adapting video image 3D denoise method |
| CN102377953A (en) * | 2011-10-17 | 2012-03-14 | 天津天地伟业数码科技有限公司 | Filtering and denoising system and filtering and denoising method for video data |
| WO2012056792A1 (en) * | 2010-10-29 | 2012-05-03 | シャープ株式会社 | Image processing device and image processing method |
| CN102497497A (en) * | 2011-12-05 | 2012-06-13 | 四川九洲电器集团有限责任公司 | Method for dynamically adjusting threshold in image denoising algorithm |
| CN102629970A (en) * | 2012-03-31 | 2012-08-08 | 广东威创视讯科技股份有限公司 | Denoising method and system for video images |
| CN102968773A (en) * | 2011-09-01 | 2013-03-13 | 伊姆普斯封闭式股份有限公司 | Method of image noise reduction |
| CN103369209A (en) * | 2013-07-31 | 2013-10-23 | 上海通途半导体科技有限公司 | Video noise reduction device and video noise reduction method |
| CN103428409A (en) * | 2012-05-15 | 2013-12-04 | 深圳中兴力维技术有限公司 | Video denoising processing method and device based on fixed scene |
-
2013
- 2013-12-20 CN CN201310712320.5A patent/CN103702016B/en not_active Expired - Fee Related
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101001317A (en) * | 2007-01-22 | 2007-07-18 | 华为技术有限公司 | Method and device of vedio noise reduction |
| CN101448077A (en) * | 2008-12-26 | 2009-06-03 | 四川虹微技术有限公司 | Self-adapting video image 3D denoise method |
| WO2012056792A1 (en) * | 2010-10-29 | 2012-05-03 | シャープ株式会社 | Image processing device and image processing method |
| CN102968773A (en) * | 2011-09-01 | 2013-03-13 | 伊姆普斯封闭式股份有限公司 | Method of image noise reduction |
| CN102377953A (en) * | 2011-10-17 | 2012-03-14 | 天津天地伟业数码科技有限公司 | Filtering and denoising system and filtering and denoising method for video data |
| CN102497497A (en) * | 2011-12-05 | 2012-06-13 | 四川九洲电器集团有限责任公司 | Method for dynamically adjusting threshold in image denoising algorithm |
| CN102629970A (en) * | 2012-03-31 | 2012-08-08 | 广东威创视讯科技股份有限公司 | Denoising method and system for video images |
| CN103428409A (en) * | 2012-05-15 | 2013-12-04 | 深圳中兴力维技术有限公司 | Video denoising processing method and device based on fixed scene |
| CN103369209A (en) * | 2013-07-31 | 2013-10-23 | 上海通途半导体科技有限公司 | Video noise reduction device and video noise reduction method |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112805990A (en) * | 2018-11-15 | 2021-05-14 | 深圳市欢太科技有限公司 | Video processing method and device, electronic equipment and computer readable storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103702016B (en) | 2017-06-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102769722B (en) | Time-space domain hybrid video noise reduction device and method | |
| US7657113B2 (en) | Auto-regressive method and filter for denoising images and videos | |
| CN107403413B (en) | Video multi-frame denoising and enhancing method | |
| Lie et al. | Motion vector recovery for video error concealment by using iterative dynamic-programming optimization | |
| US9813708B2 (en) | Method and system of coding prediction for screen video | |
| US7940333B2 (en) | Gradation control apparatus and gradation control method | |
| Jamali et al. | Fast HEVC intra mode decision based on edge detection and SATD costs classification | |
| KR20070116717A (en) | Method and device for measuring mpeg noise strength of compressed digital image | |
| US20110129020A1 (en) | Method and apparatus for banding artifact detection | |
| JP2010119121A (en) | Method and apparatus for coding image | |
| JP2012165365A (en) | Digital image stabilization method with adaptive filtering | |
| EP2321796B1 (en) | Method and apparatus for detecting dark noise artifacts | |
| CN101969568A (en) | Frame rate up conversion-oriented motion estimation method | |
| CN102509311A (en) | Motion detection method and device | |
| Dai et al. | Color video denoising based on combined interframe and intercolor prediction | |
| CN103581507A (en) | Method dynamically adjusting threshold value through mean square error in de-noising algorithm | |
| CN103702016A (en) | Video denoising method and device | |
| CN106878754A (en) | A kind of 3D video depths image method for choosing frame inner forecast mode | |
| US8831354B1 (en) | System and method for edge-adaptive and recursive non-linear filtering of ringing effect | |
| CN101546425B (en) | A vector filtering method of double-threshold type color image | |
| Fernandez et al. | Integrated H. 264 region-of-interest detection, tracking and compression for surveillance scenes | |
| Ma et al. | A fast background model based surveillance video coding in HEVC | |
| CN101778297A (en) | Interference elimination method of image sequence | |
| Matsuo et al. | Pre-processing equipment of 8K video codec with low-pass filtering and noise reduction functions | |
| CN102377953A (en) | Filtering and denoising system and filtering and denoising method for video data |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CP03 | Change of name, title or address |
Address after: Kezhu road high tech Industrial Development Zone, Guangzhou city of Guangdong Province, No. 233 510670 Patentee after: VTRON GROUP Co.,Ltd. Address before: 510663 Guangzhou province high tech Industrial Development Zone, Guangdong, Cai road, No. 6, No. Patentee before: VTRON TECHNOLOGIES Ltd. |
|
| CP03 | Change of name, title or address | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20170609 Termination date: 20211220 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |