CN102170519A - High-efficient image denoising algorithm and hardware realization apparatus of algorithm - Google Patents
High-efficient image denoising algorithm and hardware realization apparatus of algorithm Download PDFInfo
- Publication number
- CN102170519A CN102170519A CN2010102068138A CN201010206813A CN102170519A CN 102170519 A CN102170519 A CN 102170519A CN 2010102068138 A CN2010102068138 A CN 2010102068138A CN 201010206813 A CN201010206813 A CN 201010206813A CN 102170519 A CN102170519 A CN 102170519A
- Authority
- CN
- China
- Prior art keywords
- noise
- filtering
- algorithm
- salt
- image
- 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.)
- Pending
Links
- 238000001914 filtration Methods 0.000 claims abstract description 67
- 239000006002 Pepper Substances 0.000 claims abstract description 30
- 238000001514 detection method Methods 0.000 claims abstract description 12
- 230000000694 effects Effects 0.000 claims abstract description 9
- 238000000034 method Methods 0.000 claims description 9
- 238000012545 processing Methods 0.000 claims description 8
- 238000009792 diffusion process Methods 0.000 claims description 4
- 238000013459 approach Methods 0.000 claims description 2
- 238000012423 maintenance Methods 0.000 claims description 2
- 238000004321 preservation Methods 0.000 claims description 2
- 239000011159 matrix material Substances 0.000 abstract description 5
- 238000004458 analytical method Methods 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000011946 reduction process Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Landscapes
- Image Processing (AREA)
- Picture Signal Circuits (AREA)
Abstract
The present invention discloses a high-efficient image denoising algorithm and a hardware realization apparatus of the algorithm. The technical scheme comprises the steps of carrying out a gradient pre-detection of a central point of a (3*3) matrix, determining whether the noise is salt-and-pepper noise by the gradient, carrying out fast median filtering if the noise is the salt-and-pepper noise and carrying out orientation average filtering to filter Gauss noise if the noise is not the salt-and-pepper noise, and directly outputting non-noise points. The technical scheme can obtain an effect exceeding the effect of only carrying out median filtering or only carrying out average filtering under the hardware realization cost of being in need of only four rows of row cache, and has a better retentivity of image edges, and has flexible selectivity in filtering mode.
Description
Technical field:
The present invention relates to a kind of video image processing system, more particularly, when relating to video image and handling, a kind of image is carried out noise reduction algorithm and hardware implement device thereof efficiently.
Background technology:
Video signal is in the process of production, processing, transmission, inevitably to be subjected to some interference of noise, modal noise type is salt-pepper noise and Gaussian noise, have a strong impact on the visual effect of video image, therefore video image has been carried out the important step that the filtering noise reduction becomes processing system for video.
Conventional filtering algorithm is divided into linear filtering and nonlinear filtering, mean filter is the most frequently used linear filtering algorithm, can very effective filtering Gaussian noise, make that simultaneously image thickens, cause losing of image edge information, for salt-pepper noise, mean filter not only can not filtering, can cause noise diffusion to salt-pepper noise on the contrary.Medium filtering is the most frequently used nonlinear filtering algorithm, can very effective filtering noise number be lower than half salt-pepper noise of number of pixels in the filter range, also can be used as the filtering of salt-pepper noise point to the edge of image pixel, make the edge thicken, simultaneously very little effect is played in the filtering of Gaussian noise.Median filtering algorithm need carry out the ordering of a plurality of pixels, and along with the ordering that increases of pixel number need be than long operation time, the algorithm of getting intermediate value fast becomes the hard-wired key of medium filtering.
Noise is random distribution in video image, thereby so between considered frame noise reduction can on average obtain noise than the better effect of noise reduction in the frame to multiple image, but the interframe noise reduction often needs to store multiple image, needs extra frame memory space.Therefore the low hardware of a kind of needs realizes that cost can become the target of research simultaneously to the highly effective algorithm that salt-pepper noise and Gaussian noise obtain filtering noise reduction preferably.
Summary of the invention:
The object of the present invention is to provide a kind of efficient noise reduction algorithm and hardware implement device thereof that video image is handled that be used for.In particular, this method is when noise reduction process is carried out in the video image processing, and image is carried out the gradient pre-detection, selects the algorithm of Fast Median Filtering or direction mean filter according to the result who detects, and provides the hardware implement device of algorithm.
Introduce concrete technical scheme of the present invention below in detail:
A kind of algorithm of image noise reduction efficiently and hardware implement device.
This system comprises:
A kind of algorithm of image noise reduction efficiently and hardware implement device carry out the gradient pre-detection to image, and the result of gradient pre-detection is carried out analysis and judgement, select the algorithm of suitable noise reduction;
A kind of algorithm of image noise reduction efficiently and hardware implement device carry out Fast Median Filtering to salt-pepper noise.
A kind of algorithm of image noise reduction efficiently and hardware implement device are to Gaussian noise travel direction mean filter.
A kind of algorithm of image noise reduction efficiently and hardware implement device can be selected filtering algorithm flexibly or not have filtering and directly export, and hardware realizes taking the mode of concurrent operation.
It is characterized in that:
The method of the invention is carried out the gradient pre-detection to image, and the result who detects according to gradient selects suitable noise reduction algorithm.
The method of the invention is carried out medium filtering to salt-pepper noise, and computational speed is accelerated in concurrent operation.
The method of the invention avoids still undetermined salt-pepper noise to treat the influence of noise reduction pixel in the filtering Gaussian noise to Gaussian noise travel direction mean filter.
The method of the invention can be selected filtering algorithm flexibly or not have filtering and directly export, and keeps image edge information, and whole denoising device takes the mode of parallel processing to quicken arithmetic speed.
Description of drawings:
Further specify the present invention below in conjunction with the drawings and specific embodiments.
Fig. 1 is the flow chart of algorithm of the present invention.
Fig. 2 is that the gradient of algorithm of the present invention detects figure.
Fig. 3 is the Fast Median Filtering device structure chart of device of the present invention.
Fig. 4 is the direction mean filter algorithm pattern of device of the present invention.
Fig. 5 is the whole hardware implementation structure figure of device of the present invention.
Embodiment:
Further set forth the present invention below in conjunction with accompanying drawing.
Be illustrated in figure 1 as the flow chart of efficient noise reduction algorithm, at first form one 3 for an amplitude and noise acoustic image and take advantage of 3 matrix, carry out the gradient pre-detection, the result that gradient is detected carries out analysis and judgement and whether belongs to salt-pepper noise, if salt-pepper noise then enters the Fast Median Filtering device, otherwise the approach axis mean filter or directly output preserve, and select the data exported according to the result that gradient detects, obtain the image behind the noise reduction at last.
Be illustrated in figure 2 as gradient detection figure, select 3 row, 3 row to form one 3 for piece image and take advantage of 3 matrix, the Grad for the treatment of filtering pixel horizontal direction, vertical direction, 45 degree directions and 135 degree directions is calculated for treating the filtering pixel in the matrix central point, and computing formula is:
D
0＝|2e-d-f|
D
90＝|2e-b-h|
D
45＝|2e-c-g|
D
135＝|2e-a-i|
Formula 1
If the Grad of 4 directions is then thought and treated that the filtering point is a salt-pepper noise point all greater than a threshold value Th, select to carry out medium filtering; If the Grad of 4 directions greater than the number of threshold value Th less than 4, treat that then the filtering point may be edge pixel point or flat site, select directly preservation or to current some travel direction mean filter, so non-salt-pepper noise point is not carried out medium filtering, thereby reduced medium filtering level and smooth and fuzzy to the edge, well kept image edge information.
Be illustrated in figure 3 as Fast Median Filtering device structure chart, selection is to the mode of 9 somes parallel processing rather than directly to 9 somes acquisition median point that sorts, accelerated to find the speed of 9 some median points greatly, at first 9 points are divided into 3 groups, the first order is carried out the ordering of 3 points, and the maximum of 3 sorting units, intermediate value, minimum value output is sent to 2 grades, ask 3 minimum values in the maximum respectively for the 2nd grade, the maximum of the intermediate value of 3 intermediate values and 3 minimum values, and 3 results are sent into 3rd level, 3rd level is obtained the intermediate value of 3 input data, is the intermediate value of 9 points.
Be illustrated in figure 4 as direction mean filter algorithm pattern, obtain the gradient minimum value of 4 directions by gradient testing result shown in Figure 2, the direction of supposing the gradient minimum value as shown in Figure 4 is 135 degree directions, treat that then grey scale pixel value is by the mean values decision of 3 pixels of 135 degree directions after the filtering of filtering pixel, computing formula is:
e′＝(a+e+i)/3
Formula 2
Because the time to e point mean filter, f, g, h, i point are not made gradient as yet and are detected, therefore can not determine whether be the salt-pepper noise point on every side, if point then can cause the diffusion of salt-pepper noise to the mean filter that e is ordered for salt-pepper noise point on every side, therefore select the direction of the direction of gradient minimum as mean filter, reduce the baneful influence that the salt-pepper noise diffusion brings to greatest extent, also reduced the blooming of 9 mean filters of image, obtained filter effect preferably.
Be illustrated in figure 5 as the whole hardware implementation structure figure of efficient image noise reduction algorithm, be mainly 3 modules, gradient detection, Fast Median Filtering, direction mean filter, in order to improve arithmetic speed, take the mode of concurrent operation, take advantage of to 3 that 3 matrix is parallel to carry out the direction mean filter and the computing of gradient pre-detection of Fast Median Filtering, 4 directions and provide the minimal gradient direction, select the output of filtering again according to the result of gradient pre-detection, make whole filtering in a clock cycle, to finish.
Whole noise reduction filtering algorithm only needs the row cache space of one 3 row just can finish the process of filtering salt-pepper noise and Gaussian noise, and image edge information is had good maintenance effect, its to the filter effect of salt-pepper noise and Gaussian noise and hardware realize cost than be much higher than medium filtering only, mean filter, the first medium filtering noise reduction algorithm of mean filter and other two-dimensional filterings more only.
So far finished the process of filtering noise reduction.
Above-mentioned algorithm design and hardware circuit implementation structure are a kind of typical embodiment of device of the present invention, for those skilled in the art, on the basis of said apparatus, can transplant, can realize purpose of the present invention equally more complicated filtering algorithm.But this variation obviously should be in the protection range of claims of the present invention.
Claims (6)
1. image noise reduction algorithm and hardware implement device efficiently is characterized in that may further comprise the steps:
(1) image is carried out the gradient pre-detection, the result who detects according to gradient selects suitable noise reduction algorithm;
(2) salt-pepper noise is carried out medium filtering, computational speed is accelerated in concurrent operation;
(3), in the filtering Gaussian noise, avoid still undetermined salt-pepper noise to treat the influence of noise reduction pixel to Gaussian noise travel direction mean filter;
(4) select filtering algorithm or do not have filtering and directly export flexibly, keep image edge information, whole denoising device takes the mode of parallel processing to improve arithmetic speed.
2. according to described a kind of algorithm of image noise reduction efficiently of claim 1 and hardware implement device, it is characterized in that:
Adopt in the step 1 image is carried out the gradient pre-detection, the result that gradient is detected analyzes, and judges whether to belong to salt-pepper noise, if salt-pepper noise then enters the Fast Median Filtering device, otherwise approach axis mean filter or directly output preservation.
3. according to described a kind of algorithm of image noise reduction efficiently of claim 1 and hardware implement device, it is characterized in that:
Adopt in the step 2 salt-pepper noise is carried out Fast Median Filtering, 9 points are taked the mode of parallel processing rather than directly to 9 somes acquisition median point that sorts, accelerated to find the speed of 9 some median points greatly.
4. according to described a kind of algorithm of image noise reduction efficiently of claim 1 and hardware implement device, it is characterized in that:
In the step 3 to Gaussian noise travel direction mean filter, the mean value of 3 pixels of selecting gradient minimum value direction as filtering after pixel value, the noise diffusion phenomenon of avoiding still undetermined salt-pepper noise to cause in the time of the filtering Gaussian noise.
5. according to described a kind of algorithm of image noise reduction efficiently of claim 1 and hardware implement device, it is characterized in that:
Select filtering algorithm in the step 4 flexibly or do not have filtering and directly export, keep image edge information, hardware is realized taking the mode of concurrent operation to carry out, and makes whole filtering finish in a clock cycle.
6. according to described a kind of algorithm of image noise reduction efficiently of claim 1 and hardware implement device, whole noise reduction filtering algorithm only needs the row cache space of one 3 row just can finish the process of filtering salt-pepper noise and Gaussian noise, and image edge information is had good maintenance effect, its to the filter effect of salt-pepper noise and Gaussian noise and hardware realize cost than be much higher than medium filtering only, mean filter, the first medium filtering noise reduction algorithm of mean filter and other two-dimensional filterings more only.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2010102068138A CN102170519A (en) | 2010-06-22 | 2010-06-22 | High-efficient image denoising algorithm and hardware realization apparatus of algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2010102068138A CN102170519A (en) | 2010-06-22 | 2010-06-22 | High-efficient image denoising algorithm and hardware realization apparatus of algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN102170519A true CN102170519A (en) | 2011-08-31 |
Family
ID=44491480
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2010102068138A Pending CN102170519A (en) | 2010-06-22 | 2010-06-22 | High-efficient image denoising algorithm and hardware realization apparatus of algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102170519A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102752483A (en) * | 2012-06-12 | 2012-10-24 | 天津天地伟业数码科技有限公司 | Filtering noise reduction system and filtering noise reduction method based on FPGA (field programmable gate array) platform |
| CN103414845A (en) * | 2013-07-24 | 2013-11-27 | 中国航天科工集团第三研究院第八三五七研究所 | Self-adaptive video image noise reducing method and noise reducing system |
| CN103945091A (en) * | 2014-04-22 | 2014-07-23 | 苏州大学 | Digital image filter circuit design method based on FPGA evolutionary learning |
| CN105260991A (en) * | 2015-09-29 | 2016-01-20 | 中国电子科技集团公司第四十四研究所 | Adaptive hybrid noise reduction method for CMOS image sensor noise reduction processing |
| CN116977207A (en) * | 2023-03-07 | 2023-10-31 | 广西科技大学 | Image noise reduction method based on selective smoothing filtering |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060109930A1 (en) * | 2003-11-05 | 2006-05-25 | O'sullivan Robert | Method and device for deriving a predistorted signal |
| CN101043581A (en) * | 2006-03-21 | 2007-09-26 | 中国科学院计算技术研究所 | Image de-noise process of multi-template mixed filtering |
| CN101094312A (en) * | 2006-06-20 | 2007-12-26 | 西北工业大学 | Self-adapting method for filtering image with edge being retained |
| US20080025630A1 (en) * | 2006-07-28 | 2008-01-31 | Atmel Corporation | Adaptive noise filtering for reducing imager noise in digital image applications |
| CN101123681A (en) * | 2007-09-20 | 2008-02-13 | 宝利微电子系统控股公司 | A digital image noise reduction method and device |
| CN101431606A (en) * | 2008-12-01 | 2009-05-13 | 上海广电（集团）有限公司中央研究院 | Self-adapting denoising processing method based on edge detection |
-
2010
- 2010-06-22 CN CN2010102068138A patent/CN102170519A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060109930A1 (en) * | 2003-11-05 | 2006-05-25 | O'sullivan Robert | Method and device for deriving a predistorted signal |
| CN101043581A (en) * | 2006-03-21 | 2007-09-26 | 中国科学院计算技术研究所 | Image de-noise process of multi-template mixed filtering |
| CN101094312A (en) * | 2006-06-20 | 2007-12-26 | 西北工业大学 | Self-adapting method for filtering image with edge being retained |
| US20080025630A1 (en) * | 2006-07-28 | 2008-01-31 | Atmel Corporation | Adaptive noise filtering for reducing imager noise in digital image applications |
| CN101123681A (en) * | 2007-09-20 | 2008-02-13 | 宝利微电子系统控股公司 | A digital image noise reduction method and device |
| CN101431606A (en) * | 2008-12-01 | 2009-05-13 | 上海广电（集团）有限公司中央研究院 | Self-adapting denoising processing method based on edge detection |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102752483A (en) * | 2012-06-12 | 2012-10-24 | 天津天地伟业数码科技有限公司 | Filtering noise reduction system and filtering noise reduction method based on FPGA (field programmable gate array) platform |
| CN103414845A (en) * | 2013-07-24 | 2013-11-27 | 中国航天科工集团第三研究院第八三五七研究所 | Self-adaptive video image noise reducing method and noise reducing system |
| CN103945091A (en) * | 2014-04-22 | 2014-07-23 | 苏州大学 | Digital image filter circuit design method based on FPGA evolutionary learning |
| CN103945091B (en) * | 2014-04-22 | 2017-03-29 | 苏州大学 | A kind of digital picture filter circuit design method based on FPGA evolutionary learnings |
| CN105260991A (en) * | 2015-09-29 | 2016-01-20 | 中国电子科技集团公司第四十四研究所 | Adaptive hybrid noise reduction method for CMOS image sensor noise reduction processing |
| CN105260991B (en) * | 2015-09-29 | 2017-10-03 | 中国电子科技集团公司第四十四研究所 | ADAPTIVE MIXED noise-reduction method for cmos image sensor noise reduction process |
| CN116977207A (en) * | 2023-03-07 | 2023-10-31 | 广西科技大学 | Image noise reduction method based on selective smoothing filtering |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101087365B (en) | A method for filtering image mixed noise | |
| CN100505832C (en) | An image denoising method based on multi-template hybrid filtering | |
| CN103761707B (en) | A kind of mean filter method rapidly and efficiently removing image salt-pepper noise | |
| CN102170519A (en) | High-efficient image denoising algorithm and hardware realization apparatus of algorithm | |
| CN103414845A (en) | Self-adaptive video image noise reducing method and noise reducing system | |
| Youlian et al. | An improved median filtering algorithm combined with average filtering | |
| CN102752483A (en) | Filtering noise reduction system and filtering noise reduction method based on FPGA (field programmable gate array) platform | |
| CN103093443A (en) | Image impulse noise self-adaptive filtering method based on genetic algorithm-back propagation (GA-BP) neural network | |
| CN104899842B (en) | The adaptive extreme value median filter method of sequence for remote line-structured light image | |
| CN105590301B (en) | The Impulsive Noise Mitigation Method of adaptive just oblique diesis window mean filter | |
| CN109903254A (en) | An Improved Bilateral Filtering Method Based on Poisson Kernel | |
| CN111523550A (en) | Image identification method, system, medium and electronic device | |
| CN116503293B (en) | Method and system for removing salt and pepper noise of Bayer image | |
| CN106991753B (en) | A kind of image binaryzation method and device | |
| CN103714523A (en) | Image sharpening system and image sharpening method on basis of Kirsch operators | |
| CN118229568B (en) | Image edge de-pseudo noise reduction method and device, equipment and storage medium | |
| CN109993752A (en) | A kind of binary image generation method | |
| CN112652004A (en) | Image processing method, device, equipment and medium | |
| CN105141963A (en) | Image motion estimation method and device | |
| CN100367770C (en) | Method for removing isolated noise point in video | |
| CN110992279A (en) | A Multilevel Median Filtering Method for Smoothing and Denoising Metallographic Images | |
| Ji et al. | An effective self-adaptive mean filter for mixed noise | |
| CN107909555B (en) | Sharpness-keeping grid noise elimination method | |
| Sree et al. | Edge preserving algorithm for impulse noise removal using FPGA | |
| Kaur et al. | An improved weighted median filter for the image processing application |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C53 | Correction of patent of invention or patent application | ||
| CB02 | Change of applicant information |
Address after: 201203, 11, Lane 572, Lane 115, blue wave road, Zhangjiang hi tech park, Shanghai Applicant after: Shanghai InfoTM Microelectronics Co., Ltd. Address before: 201203, 11, Lane 572, Lane 115, blue wave road, Zhangjiang hi tech park, Shanghai Applicant before: Shanghai InfoTM Microelectronics Co., Ltd. |
|
| COR | Change of bibliographic data |
Free format text: CORRECT: APPLICANT; FROM: SHANGHAI INFOT MICROELECTRONICS CO., LTD. TO: SHANGHAI INFOTM MICROELECTRONICS CO., LTD. |
|
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20110831 |