CN101853489A - Video image denoising device and method - Google Patents
Video image denoising device and method Download PDFInfo
- Publication number
- CN101853489A CN101853489A CN200910106287A CN200910106287A CN101853489A CN 101853489 A CN101853489 A CN 101853489A CN 200910106287 A CN200910106287 A CN 200910106287A CN 200910106287 A CN200910106287 A CN 200910106287A CN 101853489 A CN101853489 A CN 101853489A
- Authority
- CN
- China
- Prior art keywords
- noise reduction
- mixing constant
- noise
- image
- image data
- 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
- Image Processing (AREA)
- Facsimile Image Signal Circuits (AREA)
- Picture Signal Circuits (AREA)
Abstract
The invention discloses a video image denoising device. The device comprises an original image inputting unit, a denoising unit, a mixing coefficient judging device and a mixer, wherein the original image inputting unit inputs original image data to the denoising unit and simultaneously to the mixing coefficient judging device and the mixer; and the denoised image data output by the denoising unit, the mixing coefficient output by the mixing coefficient judging device and the original image data are mixed in the mixer to acquire output image data. The invention also discloses a video image denoising method. The device and the method have the advantages of removing noise in an image, keeping original details of the image, and simultaneously avoiding the transition problem of uniform images.
Description
Technical field
The present invention relates to a kind of video image processing device and method, specifically relate to a kind of device and method that video image is carried out noise reduction process.
Background technology
The image noise reduction technology is widely used in the video image processing, and the purpose of image noise reduction is to try one's best and restores original image undistortedly, and noise removing.
Relatively the common noise signal is Gaussian noise and impulsive noise.Impulsive noise is the isolated noise of sneaking into image, and noise spot generally can be used medium filtering filtering this noise with the spot correlation degree is very poor on every side.Gaussian noise is the uniform white noise of sneaking in the image, and irrelevant with image itself, the method for this noise of filtering generally adopts one dimension or two-dimentional average filter, one dimension or 2-d gaussian filters, perhaps one dimension or two-dimentional weighted filtering etc.
Yet the elimination degree of noise and the distortion level of image are mutual restriction, and noise removing must be many more, often image information also lose many more, corresponding image is fuzzy more.
Gaussian filtering is equivalent to image is made smoothing processing, method is the neighborhood of noise spot and noise spot one dimension or two dimension to be pressed fixed coefficient do weighted mean, and smoothness can be passed through a parameter regulation, and image is level and smooth more, the noise removing effect is good more, and respective image is fuzzy more.Average filter belongs to a kind of limiting case of Gaussian noise, and its smooth effect is the strongest, and it is very fuzzy that image also becomes.
In vedio noise reduction, apply to temporal correlation toward contact, utilize the time to go up adjacent picture frame and make smoothing processing with filtering noise.This method generally also is that Gaussian noise is worked, and therefore has the same characteristic of above-mentioned gaussian filtering.
In addition, whether thresholding selects one dimension or two-dimentional weighted filtering big by the difference of judging noise spot and its one dimension or two-dimentional neighborhood point, judges whether these neighborhood points are used for doing weighted mean.This method can be suitable balance denoising performance and image blur problem, but, because the neighborhood operation point that each point is chosen in the image is all inequality, tend to cause image transition inhomogeneous, especially transform to the sudden change zone from mild zone at image, for example, the profile place of people's face or blue sky and white cloud intersection, this noise reduction mode can cause and occur the inhomogeneous and piecemeal phenomenon of tangible transition in the image.
Therefore, at present, need find a kind of good noise reduction mode, can remove noise to greatest extent, can well keep the original details of image again, simultaneously, can not cause the transition problem of even image again.
Summary of the invention
For solving the technical matters of above-mentioned noise reduction degree and the certain contradiction of image blurring existence, the present invention proposes the denoising device that a kind of video image is handled, comprise original image input block, noise reduction unit, also comprise mixing constant decision device, mixer, wherein, the original image input block inputs to noise reduction unit with raw image data, input to mixing constant decision device and mixer simultaneously, the mixing constant of the view data behind the noise reduction unit output noise reduction, the output of mixing constant decision device and raw image data mix in mixer and obtain output image data.
Described noise reduction unit adopts one dimension template noise reduction.
Described noise reduction unit adopts the n * n template noise reduction of two dimension, and wherein, n is an odd number.
A kind of video image noise reducing method, this method comprises the steps:
Step S200: raw image data is inputed in noise reduction unit, mixing constant decision device and the mixer;
Step S201: noise reduction unit obtains view data behind the noise reduction through the noise reduction template successively with original image;
Step S202: the mixing constant decision device is judged the local edge of current point according to the raw image data of input, calculates mixing constant k1 and k2 according to this local edge;
Step S203: with the view data that raw image data and noise reduction unit obtain,, in mixer, mix, obtain output image data according to mixing constant k1 and k2.
When Gaussian noise was carried out noise reduction, the computing method of mixing constant k1 and k2 comprised the steps: among the described step S202
Step S400: input original image;
Step S401: ask current point and the interior pixel difference value of certain neighborhood, wherein the light tone component of current pixel point respectively with this neighborhood in the respective component of each pixel subtract each other, obtain several difference value, the number of this each component difference value is by the template decision of noise reduction;
Step S402: with the difference maximization, the difference value on each component that obtains is compared, obtain maximum different value wherein;
Step S403: gain control, maximum different value and the gain factor that obtains among the step S402 multiplied each other, the decision original image participates in the degree of calculating;
Step S404: the generation of mixing constant, the difference value amplitude limit to 0 that obtains after gain factor multiplies each other among the step S403 is arrived between the system data maximal value, and normalize to 0～1, and described normalized value is mixing constant k1, mixing constant k2 and mixing constant k1 sum are 1.
Gain factor can dispose in real time by register among the described step S403.
When the paired pulses noise carried out noise reduction, the computing method of mixing constant k1 and k2 comprised the steps: among the described step S202
Step S500: input original image;
Step S501: ask current point and the interior pixel difference value of certain neighborhood, wherein the light tone component of current pixel point respectively with this neighborhood in the respective component of each pixel subtract each other, obtain several difference value, the number of this each component difference value is by the template decision of noise reduction;
Step S502: difference is minimized, the difference value on resulting each component is compared, obtain minimum value diff_min wherein;
Step S503: the threshold value amplitude limit, if the difference minimum value diff_min that obtains among a default threshold epsilon and the step S502 relatively greater than this threshold epsilon, thinks that then this point is a noise spot; Otherwise, judge that this point is the image border point;
Step S504: as described difference minimum value diff_min during less than described threshold epsilon, then the computing method of mixing constant k1 are shown below:
K1＝1-diff_min/ε，
k2＝diff_min/ε；
As described difference minimum value diff_min during greater than described threshold epsilon, then mixing constant k1 is 1, and k2 is 0.
Threshold epsilon can dispose in real time by register among the described step S503.
The concrete grammar that mixes among the described step S203 is for to multiply each other mixing constant k1 and original image, and the data behind k2 and the noise reduction multiply each other, and again two product additions is obtained described output image data.
The method realization that the denoising device of a kind of video image of the present invention and method adopt former figure and mix through the image overlay of traditional filtering noise reduction output, can remove the noise in the image, can well keep the original details of image again, simultaneously, can not cause the transition problem of even image again, wherein, hybrid cytokine is by the decision of the architectural feature of image itself, if edge region is selected the large percentage of former figure; At flat site, then select the large percentage of noise reduction image, thereby can solve the problem that the image border fogs in noise reduction process, restore original image as far as possible undistortedly, and noise removing.
Description of drawings
Fig. 1 is the structured flowchart of traditional denoising device;
Fig. 2 is the structured flowchart of specific embodiment of the invention video image denoising device;
Fig. 3 is the structured flowchart of a kind of embodiment of the present invention;
Fig. 4 is the process flow diagram of specific embodiment of the invention mixing constant decision device during to the Gaussian noise noise reduction;
The process flow diagram of mixing constant decision device when Fig. 5 is specific embodiment of the invention paired pulses noise noise reduction.
Embodiment
Below in conjunction with accompanying drawing the specific embodiment of the invention is specified.
Be illustrated in figure 1 as the structured flowchart of traditional denoising device, this structure comprises original image input block, noise reduction unit, and the original image input block is exported raw image data to noise reduction unit, output image data behind the noise reduction unit noise reduction.Traditional denoising device is just output after noise reduction process, can cause the fuzzy of image when noise reduction is outstanding.
Be illustrated in figure 2 as the structured flowchart of a kind of video image denoising device of the specific embodiment of the invention, this structure comprises original image input block 101, mixing constant decision device 102, noise reduction unit 103, mixer 104, original image input block 101 inputs to noise reduction unit 103 with raw image data, input to mixing constant decision device 102 and mixer 104 simultaneously, the view data behind the noise reduction unit 103 output noise reductions, the coefficient k 1 of mixing constant decision device 102 outputs and k2 and raw image data mix in mixer 104 and obtain output image data.
Wherein, described noise reduction unit 103 adopts one dimension or two-dimentional noise reduction template to carry out noise reduction, and this noise reduction template can be selected corresponding template of the prior art according to noise type, as average filter, gaussian filtering, medium filtering etc., and template can be one dimension, also can be the n * n of two dimension, wherein n is an odd number.
Described mixer 104 mixes the picture signal behind original image signal and the noise reduction according to corresponding coefficient, wherein coefficient is calculated by mixing constant decision device 102, according to the difference of noise type, the difference of filtering template, the method for calculating mixing constant k1, k2 is also different.
According to image noise reduction apparatus as shown in Figure 2, image denoising method of the present invention can be divided into following steps:
Step S200: raw image data is inputed in noise reduction unit 103, mixing constant decision device 102 and the mixer 104;
Step S201: noise reduction unit 103 obtains view data behind the noise reduction through the noise reduction template successively with original image;
Step S202: mixing constant decision device 102 is judged the local edge of current point according to the raw image data of input, calculates mixing constant k1 and k2 according to local edge;
Step S203: with the view data that raw image data and noise reduction unit 103 obtain, according to mixing constant k1 and k2, the mixing that superposes in mixer 104 obtains output image data.
Wherein, the view data that mixing constant k1 and corresponding respectively raw image data of k2 and noise reduction unit 103 obtain among the step S203, the view data that be that mixing constant k1 and raw image data weighting are multiplied each other, mixing constant k2 and noise reduction unit 103 obtains multiplies each other, and two product additions that obtain obtain mixed view data.
The computing method of mixing constant are according to the edge of image feature among the step S202, therefore any method that can be used in judgement edge or image correlation can be used for the basis for estimation of mixing constant decision device 102, as traditional sobel operator, prewitt operator, robert operator etc.When current point was close to the image border more, then the ratio of original image stack mixing was big; When current point was close to noise spot more, the ratio that the view data stack mixes behind then traditional noise reduction was big more.In the specific embodiment of the invention,, two kinds of mixing constant computing method have been proposed respectively at the noise reduction of Gaussian noise and impulsive noise.
As shown in Figure 4, when Gaussian noise is carried out noise reduction, the process flow diagram of mixing constant decision device 102, concrete steps are as follows:
Step S400: input original image;
Step S401: ask current point and the interior pixel difference value of certain neighborhood, wherein the light tone component of current pixel point respectively with this neighborhood in the respective component of each pixel subtract each other, obtain several difference value, the number of this each component difference value is by the template decision of noise reduction, when template was n * n, then the number of this each component difference value was n * n; When template was the n of one dimension, then the number of this each component difference value was n, and wherein n is natural number, and is as follows.
Step S402: with the difference maximization, the difference value on each component that obtains is compared, obtain maximum different value wherein.
Step S403: gain control, will obtain maximum different value among the step S402 and gain factor β multiplies each other, this gain factor β can dispose in real time by register, and the decision original image participates in the degree calculated.
Step S404: the generation of mixing constant is an example with 8 systems, with difference value amplitude limit to 0～255 that obtain after gain factor β multiplies each other among the step S403, and normalizes to 0～1, obtains mixing constant k1; When system is the data of other figure places, then adopt identical method to normalize to 0～1.
When noticeable, the figure place of system of the present invention is not limited to described 8 of this embodiment, when system data is the m position, then with above-mentioned difference value amplitude limit to the 0～2m-1 that obtains after gain factor β multiplies each other, renormalization to 0～1.
This mixing constant k1 is big more, illustrates that then current point and difference on every side are big more, may be a marginal point more; When described mixing constant k1 is more little, illustrate that current point and difference on every side are more little, may be the point in the smooth region more.
Mixing constant k2 is: k2=1-k1.
Simultaneously, when the paired pulses noise carries out noise reduction, then generally adopt medium filtering, corresponding mixing constant decision device 102 can adopt following workflow, and as shown in Figure 5, concrete step is as follows:
Step S500: input original image;
Step S501: ask current point and the interior pixel difference value of certain neighborhood, wherein the light tone component of current pixel point respectively with this neighborhood in the respective component of each pixel subtract each other, obtain several difference value, the number of this each component difference value is by the template decision of noise reduction, when template was n * n, then the number of this each component difference value was n * n; When template was the n of one dimension, then the number of this each component difference value was n.
Step S502: difference is minimized, the difference value on resulting each component is compared, obtain minimum value diff_min wherein.
Step S503: the threshold value amplitude limit, impulsive noise and difference on every side are generally bigger, have the pseudo-edge characteristic, therefore, if the difference minimum value diff_min that obtains among a default threshold epsilon and the step S502 relatively greater than this threshold epsilon, thinks that then this point is a noise spot; Otherwise, judge that this point is the image border point.
Wherein, the user can be by the value of the real-time configured threshold ε of register.
Step S504: the generation of mixing constant:
As described difference minimum value diff_min during less than described threshold epsilon, then the computing method of mixing constant k1 are as shown in Equation (1):
K1＝1-diff_min/ε (1)
As described difference minimum value diff_min during greater than described threshold epsilon, then mixing constant k1 is 1.
Under above-mentioned two kinds of situations, mixing constant k2 all satisfies k1 and the k2 sum is 1, i.e. k2=1-k1.
Be illustrated in figure 3 as the structured flowchart of a kind of embodiment of the present invention when dimension n is 3 in the noise reduction template, when original image input block 201 is imported data successively to mixing constant decision device 202 and noise reduction unit 203 and mixer 204, when the current point of input is the e point, because the noise reduction template is 3 * 3, then importing data is the center with the e point, 3 * 3 original image zone and 203 weightings of noise reduction template obtain the view data behind the noise reduction; Simultaneously, three component Y, U of input signal, V obtain the edge feature that current some e ordered through mixing constant decision device 202, the mixing constant k1 and the k2 that calculate are with view data behind original image input block 201 and the noise reduction weighting summation view data that obtains exporting respectively.
The present invention is not limited to described 3 * 3 noise reduction templates of this embodiment, and when the noise reduction template was n * n, then the zone that is the n * n at center with current point in the original image input block 201 participated in the calculating of current point.
Above content be in conjunction with concrete preferred implementation to further describing that the present invention did, can not assert that concrete enforcement of the present invention is confined to these explanations.For the general technical staff of the technical field of the invention, without departing from the inventive concept of the premise, can also make some simple deduction or replace, all should be considered as belonging to protection scope of the present invention.
Claims (9)
1. video image denoising device, comprise original image input block, noise reduction unit, it is characterized in that, this device also comprises mixing constant decision device, mixer, wherein, the original image input block inputs to noise reduction unit with raw image data, inputs to mixing constant decision device and mixer simultaneously, and the mixing constant of the view data behind the noise reduction unit output noise reduction, the output of mixing constant decision device and raw image data mix in mixer and obtain output image data.
2. a kind of video image denoising device according to claim 1 is characterized in that, described noise reduction unit adopts one dimension template noise reduction.
3. a kind of video image denoising device according to claim 1 is characterized in that, described noise reduction unit adopts the n * n template noise reduction of two dimension, and wherein, n is an odd number.
4. a video image noise reducing method is characterized in that, this method comprises the steps:
Step S200: raw image data is inputed in noise reduction unit, mixing constant decision device and the mixer;
Step S201: noise reduction unit obtains view data behind the noise reduction through the noise reduction template successively with original image;
Step S202: the mixing constant decision device is judged the local edge of current point according to the raw image data of input, calculates mixing constant k1 and k2 according to this local edge;
Step S203: with the view data that raw image data and noise reduction unit obtain,, in mixer, mix, obtain output image data according to mixing constant k1 and k2.
5. a kind of video image noise reducing method according to claim 4 is characterized in that, when Gaussian noise was carried out noise reduction, the computing method of k1 of mixing constant described in the step S202 and k2 comprised the steps:
Step S400: input original image;
Step S401: ask current point and the interior pixel difference value of certain neighborhood, wherein the light tone component of current pixel point respectively with this neighborhood in the respective component of each pixel subtract each other, obtain several difference value, the number of this each component difference value is by the template decision of noise reduction;
Step S402: with the difference maximization, the difference value on each component that obtains is compared, obtain maximum different value wherein;
Step S403: gain control, maximum different value and the gain factor that obtains among the step S402 multiplied each other, the decision original image participates in the degree of calculating;
Step S404: the generation of mixing constant, the difference value amplitude limit to 0 that obtains after gain factor multiplies each other among the step S403 is arrived between the system data maximal value, and normalize to 0～1, and described normalized value is mixing constant k1, mixing constant k2 and mixing constant k1 sum are 1.
6. a kind of video image noise reducing method according to claim 5 is characterized in that gain factor described in the step S403 can dispose in real time by register.
7. a kind of video image noise reducing method according to claim 4 is characterized in that, when the paired pulses noise carried out noise reduction, the computing method of k1 of mixing constant described in the step S202 and k2 comprised the steps:
Step S500: input original image;
Step S501: ask current point and the interior pixel difference value of certain neighborhood, wherein the light tone component of current pixel point respectively with this neighborhood in the respective component of each pixel subtract each other, obtain several difference value, the number of this each component difference value is by the template decision of noise reduction;
Step S502: difference is minimized, the difference value on resulting each component is compared, obtain minimum value diff_min wherein;
Step S503: the threshold value amplitude limit, if the difference minimum value diff_min that obtains among a default threshold epsilon and the step S502 relatively greater than this threshold epsilon, thinks that then this point is a noise spot; Otherwise, judge that this point is the image border point;
Step S504: as described difference minimum value diff_min during less than described threshold epsilon, then the computing method of mixing constant k1 are shown below:
K1＝1-diff_min/ε，
k2＝diff_min/ε；
As described difference minimum value diff_min during greater than described threshold epsilon, then mixing constant k1 is 1, and k2 is 0.
8. a kind of video image noise reducing method according to claim 7 is characterized in that threshold epsilon described in the step S503 can dispose in real time by register.
9. a kind of video image noise reducing method according to claim 4, it is characterized in that, the concrete grammar that mixes described in the step S203 is for to multiply each other mixing constant k1 and original image, and the data behind k2 and the noise reduction multiply each other, and again two product additions is obtained described output image data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200910106287.5A CN101853489B (en) | 2009-04-02 | 2009-04-02 | Video image denoising device and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200910106287.5A CN101853489B (en) | 2009-04-02 | 2009-04-02 | Video image denoising device and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101853489A true CN101853489A (en) | 2010-10-06 |
| CN101853489B CN101853489B (en) | 2014-03-12 |
Family
ID=42804954
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN200910106287.5A Active CN101853489B (en) | 2009-04-02 | 2009-04-02 | Video image denoising device and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101853489B (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102753100A (en) * | 2010-11-01 | 2012-10-24 | 株式会社东芝 | Ultrasonic diagnostic apparatus and ultrasonic image processing apparatus |
| CN102769722A (en) * | 2012-07-20 | 2012-11-07 | 上海富瀚微电子有限公司 | Time-space domain hybrid video noise reduction device and method |
| CN102999882B (en) * | 2011-09-13 | 2016-08-24 | 深圳艾科创新微电子有限公司 | A kind of noise-reduction method based on random replacement and system |
| CN106791284A (en) * | 2017-01-17 | 2017-05-31 | 深圳市维海德技术股份有限公司 | A kind of method and device for removing impulsive noise |
| CN108937950A (en) * | 2018-05-30 | 2018-12-07 | 梧州井儿铺贸易有限公司 | A kind of intelligent robot |
| CN110322436A (en) * | 2019-06-19 | 2019-10-11 | 广州金域医学检验中心有限公司 | Medical image processing method, device, storage medium and equipment |
| CN110706170A (en) * | 2019-09-26 | 2020-01-17 | 哈尔滨工业大学 | Denoising method for image of portable B-type ultrasonic diagnostic equipment |
| CN112215758A (en) * | 2019-07-11 | 2021-01-12 | 瑞昱半导体股份有限公司 | Denoising method based on signal-to-noise ratio |
| CN116523765A (en) * | 2023-03-13 | 2023-08-01 | 湖南兴芯微电子科技有限公司 | Real-time video image noise reduction method, device and memory |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6985636B1 (en) * | 1998-09-03 | 2006-01-10 | Semenchenko Michail Grigorievi | Image processing method |
| WO2008010375A1 (en) * | 2006-07-20 | 2008-01-24 | Hitachi Medical Corporation | Ultrasonographic device |
| CN101123681A (en) * | 2007-09-20 | 2008-02-13 | 宝利微电子***控股公司 | A digital image noise reduction method and device |
-
2009
- 2009-04-02 CN CN200910106287.5A patent/CN101853489B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6985636B1 (en) * | 1998-09-03 | 2006-01-10 | Semenchenko Michail Grigorievi | Image processing method |
| WO2008010375A1 (en) * | 2006-07-20 | 2008-01-24 | Hitachi Medical Corporation | Ultrasonographic device |
| CN101123681A (en) * | 2007-09-20 | 2008-02-13 | 宝利微电子***控股公司 | A digital image noise reduction method and device |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102753100A (en) * | 2010-11-01 | 2012-10-24 | 株式会社东芝 | Ultrasonic diagnostic apparatus and ultrasonic image processing apparatus |
| CN102753100B (en) * | 2010-11-01 | 2015-10-07 | 株式会社东芝 | Diagnostic ultrasound equipment and Ultrasonographic device |
| CN102999882B (en) * | 2011-09-13 | 2016-08-24 | 深圳艾科创新微电子有限公司 | A kind of noise-reduction method based on random replacement and system |
| CN102769722A (en) * | 2012-07-20 | 2012-11-07 | 上海富瀚微电子有限公司 | Time-space domain hybrid video noise reduction device and method |
| CN102769722B (en) * | 2012-07-20 | 2015-04-29 | 上海富瀚微电子股份有限公司 | Time-space domain hybrid video noise reduction device and method |
| CN106791284B (en) * | 2017-01-17 | 2019-11-12 | 深圳市维海德技术股份有限公司 | A kind of method and device removing impulsive noise |
| CN106791284A (en) * | 2017-01-17 | 2017-05-31 | 深圳市维海德技术股份有限公司 | A kind of method and device for removing impulsive noise |
| CN108937950A (en) * | 2018-05-30 | 2018-12-07 | 梧州井儿铺贸易有限公司 | A kind of intelligent robot |
| CN110322436A (en) * | 2019-06-19 | 2019-10-11 | 广州金域医学检验中心有限公司 | Medical image processing method, device, storage medium and equipment |
| CN110322436B (en) * | 2019-06-19 | 2020-10-02 | 广州金域医学检验中心有限公司 | Medical image processing method, device, storage medium and equipment |
| CN112215758A (en) * | 2019-07-11 | 2021-01-12 | 瑞昱半导体股份有限公司 | Denoising method based on signal-to-noise ratio |
| CN112215758B (en) * | 2019-07-11 | 2023-10-10 | 瑞昱半导体股份有限公司 | Noise removing method based on signal to noise ratio |
| CN110706170A (en) * | 2019-09-26 | 2020-01-17 | 哈尔滨工业大学 | Denoising method for image of portable B-type ultrasonic diagnostic equipment |
| CN110706170B (en) * | 2019-09-26 | 2023-03-14 | 哈尔滨工业大学 | Denoising method for image of portable B-type ultrasonic diagnostic equipment |
| CN116523765A (en) * | 2023-03-13 | 2023-08-01 | 湖南兴芯微电子科技有限公司 | Real-time video image noise reduction method, device and memory |
| CN116523765B (en) * | 2023-03-13 | 2023-09-05 | 湖南兴芯微电子科技有限公司 | Real-time video image noise reduction method, device and memory |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101853489B (en) | 2014-03-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101853489B (en) | Video image denoising device and method | |
| US11127122B2 (en) | Image enhancement method and system | |
| Li et al. | Multi-scale single image dehazing using Laplacian and Gaussian pyramids | |
| US8244054B2 (en) | Method, apparatus and integrated circuit capable of reducing image ringing noise | |
| EP2462558B1 (en) | Image processing apparatus, image processing method, and computer program | |
| CN101710415B (en) | Image enhancement coefficient adjusting method and device thereof and image enhancement method and device thereof | |
| Yüksel | Edge detection in noisy images by neuro-fuzzy processing | |
| CN107742291B (en) | Defect detection method and device for photovoltaic glass | |
| CN115631117B (en) | Image enhancement method, device, detection system and storage medium for defect detection | |
| CN103238335A (en) | Image processing device, image processing method, and program | |
| JP4731100B2 (en) | Sharpness improving method and sharpness improving device | |
| CN104091310A (en) | Image defogging method and device | |
| Habib et al. | Adaptive fuzzy inference system based directional median filter for impulse noise removal | |
| CN103886565A (en) | Nighttime color image enhancement method based on purpose optimization and histogram equalization | |
| CN104881843A (en) | Image interpolation method and image interpolation apparatus | |
| Semenishchev et al. | Computation of equidistant curve for the image with blurred contours | |
| Li et al. | An improved image defogging method based on dark channel prior | |
| CN102118547A (en) | Image weighted filtering method | |
| CN101431606A (en) | Self-adapting denoising processing method based on edge detection | |
| CN105427265A (en) | Color image contrast enhancement method and system | |
| CN113450272B (en) | Image enhancement method based on sinusoidal variation and application thereof | |
| CN103514588A (en) | Image enhancement method and system | |
| CN104599238A (en) | Image processing method and device | |
| CN110264415B (en) | Image processing method for eliminating jitter blur | |
| CN103258318B (en) | A kind of image noise reduction disposal route and system |
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 | ||
| PE01 | Entry into force of the registration of the contract for pledge of patent right |
Denomination of invention: Video image denoising device and method Effective date of registration: 20150129 Granted publication date: 20140312 Pledgee: Shenzhen high tech investment and financing Company limited by guarantee Pledgor: Aike Chuangxin Microelectronic Co., Ltd. Registration number: 2015990000082 |
|
| PLDC | Enforcement, change and cancellation of contracts on pledge of patent right or utility model | ||
| CP01 | Change in the name or title of a patent holder |
Address after: 518057 Guangdong science and technology two Road Software Park, phase 1, 4, 4 building, room two, Nanshan District high tech Zone, Shenzhen, China Patentee after: Shenzhen Shenyang electronic Limited by Share Ltd Address before: 518057 Guangdong science and technology two Road Software Park, phase 1, 4, 4 building, room two, Nanshan District high tech Zone, Shenzhen, China Patentee before: Aike Chuangxin Microelectronic Co., Ltd. |
|
| CP01 | Change in the name or title of a patent holder | ||
| PC01 | Cancellation of the registration of the contract for pledge of patent right |
Date of cancellation: 20210705 Granted publication date: 20140312 Pledgee: Shenzhen high tech investment and financing Company limited by guarantee Pledgor: ARK PIONEER MICROELECTRONICS (SHENZHEN) Co.,Ltd. Registration number: 2015990000082 |
|
| PC01 | Cancellation of the registration of the contract for pledge of patent right |