CN103559685A - Image filtering algorithm based on wavelet transformation - Google Patents
Image filtering algorithm based on wavelet transformation Download PDFInfo
- Publication number
- CN103559685A CN103559685A CN201310481368.XA CN201310481368A CN103559685A CN 103559685 A CN103559685 A CN 103559685A CN 201310481368 A CN201310481368 A CN 201310481368A CN 103559685 A CN103559685 A CN 103559685A
- Authority
- CN
- China
- Prior art keywords
- data
- threshold
- value
- assignment
- work
- 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
Landscapes
- Image Processing (AREA)
Abstract
The invention mainly relates to image filtering in digital image processing, and aims to achieve the good filtering effect and the goals of effectively reducing or removing noise in images and improving the visual quality of the images after the images are filtered. According to the technical scheme, an image filtering algorithm based on wavelet transformation includes the following steps that image processing is conducted on four adjacent pixels, the values of the pixels are respectively U, V, W and X, and DATA (1) is obtained by dividing 2 from the sum of U, V, W and X as shown in the formula DATA (1)= (U+ V+ W+X)/2; DATA (2) is obtained by dividing 2 from the value obtained by subtracting the sum of V and X from the sum of U and W, and DATA (3) is obtained by dividing 2 from the value obtained by subtracting the sum of W and X from the sum of U and V; DATA (4) is obtained by dividing 2 from the value obtained by subtracting the sum of V and W from the sum of U and X. The image filtering algorithm is mainly applied to image filtering.
Description
Technical field
The present invention relates to relate generally to the image filtering in Digital Image Processing, refer more particularly to the processing of digital picture in monitoring and the communications field.Specifically relate to the Image filter arithmetic based on wavelet transformation.
Technical background
Image is that the mankind obtain the main source with exchange message, and therefore, the application that image is processed must relate to the every aspect of human lives and work.Along with the continuous expansion of mankind's activity scope, the application that image is processed also will constantly expand thereupon.The application of Digital Image Processing is varied, and except communication, monitoring, digital picture is also applied to the fields such as medical science and space item, and image filtering is the important component part in image processing.
Summary of the invention
For overcoming the deficiencies in the prior art, the present invention is intended to obtain good filter effect, realize image after filtering after, can effectively weaken or removal of images in noise, improve the visual sense quality of image, for this reason, the technical solution used in the present invention is, Image filter arithmetic based on wavelet transformation, comprises the following steps: four adjacent pixels to carry out image processing, and the value of pixel is respectively U, V, W and X, U, V, W and X are added, and divided by 2, obtain data DATA[1], be formulated as DATA[1]=(U+V+W+X)/2; U and W are added, deduct V and X, and divided by 2, obtain DATA[2], be formulated as DATA[2]=(U-V+W-X)/2; U and V are added, deduct W and X, and divided by 2, obtain DATA[3], be formulated as DATA[3]=(U+V-W-X)/2; U and X are added, deduct V and W, and divided by 2, obtain DATA[4], be formulated as DATA[4]=(U-V-W+X)/2;
DATA_CO[1] be assigned DATA[1]; Work as DATA[2] value be less than threshold value Threshold, and during be greater than-Threshold, DATA_CO[2] assignment is zero; Work as DATA[2] value be greater than threshold value Threshold, DATA_CO[2] assignment is DATA[2]-Threshold; Work as DATA[2] value be less than threshold value-Threshold, DATA_CO[2] assignment is DATA[2]+Threshold; Work as DATA[3] value be less than threshold value Threshold, and during be greater than-Threshold, DATA_CO[3] assignment is zero; Work as DATA[3] value be greater than threshold value Threshold, DATA_CO[3] assignment is DATA[3]-Threshold; Work as DATA[3] value be less than threshold value-Threshold, DATA_CO[3] assignment is DATA[3]+Threshold; Work as DATA[4] value be less than threshold value Threshold, and during be greater than-Threshold, DATA_CO[4] assignment is zero; Work as DATA[4] value be greater than threshold value Threshold, DATA_CO[4] assignment is DATA[4]-Threshold; Work as DATA[4] value be less than threshold value-Threshold, DATA_CO[4] assignment is DATA[4]+Threshold;
To DATA_CO[1], DATA_CO[2], DATA_CO[3], DATA_CO[4] four numerical value process: to DATA_CO[1], DATA_CO[2], DATA_CO[3] and DATA_CO[4] be added, and obtain data U1 divided by 2, and be formulated as U1=(DATA_CO[1]+DATA_CO[2]+DATA_CO[3]+DATA_CO[4])/2; To DATA_CO[1] and DATA_CO[3] be added, deduct DATA_CO[2] and DATA_CO[4], and divided by 2, obtain V1, and be formulated as V1=(DATA_CO[1]-DATA_CO[2]+DATA_CO[3]-DATA_CO[4])/2; To DATA_CO[1] and DATA_CO[2] be added, deduct DATA_CO[3] and DATA_CO[4], and divided by 2, obtain W1, and be formulated as W1=(DATA_CO[1]+DATA_CO[2]-DATA_CO[3]-DATA_CO[4])/2; To DATA_CO[1] and DATA_CO[4] be added, deduct DATA_CO[2] and DATA_CO[3], and divided by 2, obtain X1, and be formulated as X1=DATA_CO[1]-DATA_CO[2]-DATA_CO[3]+DATA_CO[4])/2; U1, V1, W1, X1 is the pixel value after processing, and corresponds respectively to the position of original pixel U, V, W and X; Each pixel to piece image is processed successively and can be obtained filtered entire image according to above-mentioned algorithm.
The position of U, V, W and X is: U is directly over W, and V is in U front-right, and X is in W front-right, and X is under V.For the larger image of noise, same piece image is carried out to twice ground filtering and can reach good effect.
The present invention possesses following technique effect:
The algorithm proposing can reduce the noise component in image effectively, makes image become more clear.Not only be practically applicable to the processing of general picture, simultaneously can be for the processing of video image.
Embodiment
The processing of piece image is normally processed each pixel in image.In digital picture, each pixel is not independently, and its correlativity is large.On image frame, often there are a lot of pixels to have identical or approaching gray scale.The Image filter arithmetic that the present invention proposes is also the correlativity based on adjacent image.In order to obtain good filter effect, need to carry out image processing to four adjacent pixels, the value of pixel is respectively U, V, W and X.U, V, W and X are added, and divided by 2, obtain data DATA[1], be formulated as DATA[1]=(U+V+W+X)/2.U and W are added, deduct V and X, and divided by 2, obtain DATA[2], be formulated as DATA[2]=(U-V+W-X)/2.U and V are added, deduct W and X, and divided by 2, obtain DATA[3], be formulated as DATA[3]=(U+V-W-X)/2.U and X are added, deduct V and W, and divided by 2, obtain DATA[4], be formulated as DATA[4]=(U-V-W+X)/2.
DATA_CO[1] be assigned DATA[1].Work as DATA[2] value be less than threshold value Threshold, and during be greater than-Threshold, DATA_CO[2] assignment is zero; Work as DATA[2] value be greater than threshold value Threshold, DATA_CO[2] assignment is DATA[2]-Threshold; Work as DATA[2] value be less than threshold value-Threshold, DATA_CO[2] assignment is DATA[2]+Threshold.Work as DATA[3] value be less than threshold value Threshold, and during be greater than-Threshold, DATA_CO[3] assignment is zero; Work as DATA[3] value be greater than threshold value Threshold, DATA_CO[3] assignment is DATA[3]-Threshold; Work as DATA[3] value be less than threshold value-Threshold, DATA_CO[3] assignment is DATA[3]+Threshold.Work as DATA[4] value be less than threshold value Threshold, and during be greater than-Threshold, DATA_CO[4] assignment is zero; Work as DATA[4] value be greater than threshold value Threshold, DATA_CO[4] assignment is DATA[4]-Threshold; Work as DATA[4] value be less than threshold value-Threshold, DATA_CO[4] assignment is DATA[4]+Threshold.
In order to obtain final image value, need to be to DATA_CO[1], DATA_CO[2], DATA_CO[3], DATA_CO[4] four numerical value process.To DATA_CO[1], DATA_CO[2], DATA_CO[3] and DATA_CO[4] be added, and obtain data U1 divided by 2, be formulated as U1=(DATA_CO[1]+DATA_CO[2]+DATA_CO[3]+DATA_CO[4])/2.To DATA_CO[1] and DATA_CO[3] be added, deduct DATA_CO[2] and DATA_CO[4], and divided by 2, obtain V1, and be formulated as V1=(DATA_CO[1]-DATA_CO[2]+DATA_CO[3]-DATA_CO[4])/2.To DATA_CO[1] and DATA_CO[2] be added, deduct DATA_CO[3] and DATA_CO[4], and divided by 2, obtain W1, and be formulated as W1=(DATA_CO[1]+DATA_CO[2]-DATA_CO[3]-DATA_CO[4])/2.To DATA_CO[1] and DATA_CO[4] be added, deduct DATA_CO[2] and DATA_CO[3], and divided by 2, obtain X1, and be formulated as X1=DATA_CO[1]-DATA_CO[2]-DATA_CO[3]+DATA_CO[4])/2.U1, V1, W1, X1 is the pixel value after processing, and corresponds respectively to the position of original pixel U, V, W and X.Each pixel to piece image is processed successively and can be obtained filtered entire image according to above-mentioned algorithm.
In order to obtain good effect, threshold value Threshold will adjust according to the noise of real image.The position of U, V, W and X is: U is directly over W, and V is in U front-right, and X is in W front-right, and X is under V.For the larger image of noise, same piece image is carried out to twice ground filtering and can reach good effect.
Claims (2)
1. the Image filter arithmetic based on wavelet transformation, is characterized in that, comprises the following steps: four adjacent pixels to carry out image processing, and the value of pixel is respectively U, V, W and X.U, V, W and X are added, and divided by 2, obtain data DATA[1], be formulated as DATA[1]=(U+V+W+X)/2; U and W are added, deduct V and X, and divided by 2, obtain DATA[2], be formulated as DATA[2]=(U-V+W-X)/2; U and V are added, deduct W and X, and divided by 2, obtain DATA[3], be formulated as DATA[3]=(U+V-W-X)/2; U and X are added, deduct V and W, and divided by 2, obtain DATA[4], be formulated as DATA[4]=(U-V-W+X)/2;
DATA_CO[1] be assigned DATA[1]; Work as DATA[2] value be less than threshold value Threshold, and during be greater than-Threshold, DATA_CO[2] assignment is zero; Work as DATA[2] value be greater than threshold value Threshold, DATA_CO[2] assignment is DATA[2]-Threshold; Work as DATA[2] value be less than threshold value-Threshold, DATA_CO[2] assignment is DATA[2]+Threshold; Work as DATA[3] value be less than threshold value Threshold, and during be greater than-Threshold, DATA_CO[3] assignment is zero; Work as DATA[3] value be greater than threshold value Threshold, DATA_CO[3] assignment is DATA[3]-Threshold; Work as DATA[3] value be less than threshold value-Threshold, DATA_CO[3] assignment is DATA[3]+Threshold; Work as DATA[4] value be less than threshold value Threshold, and during be greater than-Threshold, DATA_CO[4] assignment is zero; Work as DATA[4] value be greater than threshold value Threshold, DATA_CO[4] assignment is DATA[4]-Threshold; Work as DATA[4] value be less than threshold value-Threshold, DATA_CO[4] assignment is DATA[4]+Threshold;
To DATA_CO[1], DATA_CO[2], DATA_CO[3], DATA_CO[4] four numerical value process: to DATA_CO[1], DATA_CO[2], DATA_CO[3] and DATA_CO[4] be added, and obtain data U1 divided by 2, and be formulated as U1=(DATA_CO[1]+DATA_CO[2]+DATA_CO[3]+DATA_CO[4])/2; To DATA_CO[1] and DATA_CO[3] be added, deduct DATA_CO[2] and DATA_CO[4], and divided by 2, obtain V1, and be formulated as V1=(DATA_CO[1]-DATA_CO[2]+DATA_CO[3]-DATA_CO[4])/2; To DATA_CO[1] and DATA_CO[2] be added, deduct DATA_CO[3] and DATA_CO[4], and divided by 2, obtain W1, and be formulated as W1=(DATA_CO[1]+DATA_CO[2]-DATA_CO[3]-DATA_CO[4])/2; To DATA_CO[1] and DATA_CO[4] be added, deduct DATA_CO[2] and DATA_CO[3], and divided by 2, obtain X1, and be formulated as X1=DATA_CO[1]-DATA_CO[2]-DATA_CO[3]+DATA_CO[4])/2; U1, V1, W1, X1 is the pixel value after processing, and corresponds respectively to the position of original pixel U, V, W and X; Each pixel to piece image is processed successively and can be obtained filtered entire image according to above-mentioned algorithm.
2. the Image filter arithmetic based on wavelet transformation as claimed in claim 1, is characterized in that, the position of U, V, W and X is: U is directly over W, and V is in U front-right, and X is in W front-right, and X is under V.For the larger image of noise, same piece image is carried out to twice ground filtering and can reach good effect.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310481368.XA CN103559685A (en) | 2013-10-14 | 2013-10-14 | Image filtering algorithm based on wavelet transformation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310481368.XA CN103559685A (en) | 2013-10-14 | 2013-10-14 | Image filtering algorithm based on wavelet transformation |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103559685A true CN103559685A (en) | 2014-02-05 |
Family
ID=50013924
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310481368.XA Pending CN103559685A (en) | 2013-10-14 | 2013-10-14 | Image filtering algorithm based on wavelet transformation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103559685A (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090154825A1 (en) * | 2007-12-14 | 2009-06-18 | Intel Corporation | Reduction filter based on smart neighbor selection and weighting (nrf-snsw) |
| CN101477680A (en) * | 2009-01-16 | 2009-07-08 | 天津大学 | Wavelet image denoising process based on sliding window adjacent region data selection |
| TW201112163A (en) * | 2009-09-21 | 2011-04-01 | Pixart Imaging Inc | Image denoising method |
| CN102236888A (en) * | 2011-07-22 | 2011-11-09 | 清华大学 | Image denoising method based on dual-tree discrete wavelet packet |
| CN102800056A (en) * | 2012-06-30 | 2012-11-28 | 浙江大学 | Neighborhood adaptive Bayes shrinkage image denoising method based on dual-tree complex wavelet domain |
| US20130028538A1 (en) * | 2011-07-29 | 2013-01-31 | Simske Steven J | Method and system for image upscaling |
| CN103177428A (en) * | 2013-03-21 | 2013-06-26 | 西安电子科技大学 | Synthetic aperture radar (SAR) image denoising method based on nonsubsampled directional wavelet transform and fusion |
-
2013
- 2013-10-14 CN CN201310481368.XA patent/CN103559685A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090154825A1 (en) * | 2007-12-14 | 2009-06-18 | Intel Corporation | Reduction filter based on smart neighbor selection and weighting (nrf-snsw) |
| CN101477680A (en) * | 2009-01-16 | 2009-07-08 | 天津大学 | Wavelet image denoising process based on sliding window adjacent region data selection |
| TW201112163A (en) * | 2009-09-21 | 2011-04-01 | Pixart Imaging Inc | Image denoising method |
| CN102236888A (en) * | 2011-07-22 | 2011-11-09 | 清华大学 | Image denoising method based on dual-tree discrete wavelet packet |
| US20130028538A1 (en) * | 2011-07-29 | 2013-01-31 | Simske Steven J | Method and system for image upscaling |
| CN102800056A (en) * | 2012-06-30 | 2012-11-28 | 浙江大学 | Neighborhood adaptive Bayes shrinkage image denoising method based on dual-tree complex wavelet domain |
| CN103177428A (en) * | 2013-03-21 | 2013-06-26 | 西安电子科技大学 | Synthetic aperture radar (SAR) image denoising method based on nonsubsampled directional wavelet transform and fusion |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106251318B (en) | Denoising device and method for sequence image | |
| EP3166070B1 (en) | Method for upscaling noisy images, and apparatus for upscaling noisy images | |
| NZ612089A (en) | Deblocking filtering | |
| CN101316321B (en) | Pattern noise removal method and device based on median filter | |
| CN106815821B (en) | Denoising method and device for near-infrared image | |
| CN102281386B (en) | Method and device for performing adaptive denoising on video image | |
| RU2015103069A (en) | DEVICE FOR PROCESSING IMAGES AND METHOD FOR PROCESSING IMAGES | |
| CN104796582B (en) | Video image denoising and Enhancement Method and device based on random injection retinex | |
| EP2664149A4 (en) | Deblocking filtering | |
| JP2009147922A5 (en) | Image processing method, display device, and computer program | |
| Gao et al. | Based on soft-threshold wavelet de-noising combining with Prewitt operator edge detection algorithm | |
| CN103888638A (en) | Time-space domain self-adaption denoising method based on guide filtering and non-local average filtering | |
| CN102497497B (en) | Method for dynamically adjusting threshold in image denoising algorithm | |
| CN100367770C (en) | Method for removing isolated noise point in video | |
| CN101504769B (en) | Self-adaptive noise intensity estimation method based on encoder frame work | |
| CN102903077A (en) | Rapid image de-blurring algorithm | |
| CN103905692A (en) | Simple 3D noise reduction algorithm base on motion detection | |
| CN103886558A (en) | Improved adaptive threshold wavelet denoising algorithm based on LoG operator | |
| CN103559685A (en) | Image filtering algorithm based on wavelet transformation | |
| CN102542533A (en) | Image blurring method and system | |
| CN103067647B (en) | A kind of video denoising method based on FPGA | |
| CN105469394B (en) | A kind of Intelligent target tracking based on complex environment | |
| CN110836649B (en) | Self-adaptive spatial composite ultrasonic imaging method | |
| CN103150710A (en) | Rapid mean filtering method for image | |
| CN105069764B (en) | A kind of image de-noising method and system based on Edge track |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| AD01 | Patent right deemed abandoned |
Effective date of abandoning: 20170419 |
|
| AD01 | Patent right deemed abandoned |