CN107330863A - A kind of image de-noising method estimated based on noise - Google Patents
A kind of image de-noising method estimated based on noise Download PDFInfo
- Publication number
- CN107330863A CN107330863A CN201710390337.1A CN201710390337A CN107330863A CN 107330863 A CN107330863 A CN 107330863A CN 201710390337 A CN201710390337 A CN 201710390337A CN 107330863 A CN107330863 A CN 107330863A
- Authority
- CN
- China
- Prior art keywords
- mrow
- msub
- pixel
- image
- denoising
- 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
- 238000000034 method Methods 0.000 title claims abstract description 25
- 230000011218 segmentation Effects 0.000 claims abstract description 10
- 238000009499 grossing Methods 0.000 claims description 4
- 229910010888 LiIn Inorganic materials 0.000 claims description 2
- 239000003550 marker Substances 0.000 claims description 2
- 239000000463 material Substances 0.000 abstract description 5
- 230000000694 effects Effects 0.000 description 4
- 230000003044 adaptive effect Effects 0.000 description 1
- 239000000654 additive Substances 0.000 description 1
- 230000000996 additive effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 238000012634 optical imaging Methods 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20021—Dividing image into blocks, subimages or windows
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20172—Image enhancement details
- G06T2207/20182—Noise reduction or smoothing in the temporal domain; Spatio-temporal filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20172—Image enhancement details
- G06T2207/20192—Edge enhancement; Edge preservation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a kind of image de-noising method estimated based on noise, some homology regions are classified as according to picture material using super-pixel segmentation, using flat information more can representative image noise pollution level priori, homology region more smooth in image is found out by foundation of image information entropy, and estimate that the noise criteria of smooth region is poor, the noise level of entire image is used as using the noise criteria difference of these smooth regions, reach the purpose of more accurate estimation noise level, so as to be modified according to noise level to non-local mean (NLM) denoising method, realization reasonably controls denoising degree according to noise level, noisy image can adaptively be handled, the NLM algorithms for making the image whole structure after denoising more traditional are significantly lifted, details is preferably remained while denoising.Overall process can realize automation and self energy, without manual intervention.
Description
Technical field
The present invention relates to the image de-noising method of image processing field, more particularly to a kind of estimation of noise.
Background technology
With the fast development of digital picture and multimedia technology, various types of optical imaging systems are also more and more,
People require also more and more higher, but in the transmitting procedure of image to the image quality of imaging system, can be made an uproar unavoidably by various
The pollution of sound, so as to cause acquired image to be decreased with original picture quality, not only can imaged image effect, make an uproar
The readability of image even can be influenceed when sound is serious, mistake occurs so as to cause subsequently to obtain information.Therefore, led in image procossing
In domain, it is very significant to carry out denoising to digital picture.
A kind of current conventional denoising method is excellent performance and the stronger non-local mean of edge holding capacity (NLM) figure
As Denoising Algorithm.Its basic thought is the pixel similar to treating denoising pixel to be found in entire image and with these similar pixels
Weight average value as its denoising result, the corresponding weights of the higher pixel of similarity are also higher.It the advantage is that introducing is non-
The concept of local thought and similarity based on image block.In natural image, the corresponding similar pixel of each pixel is often not
It is only limitted to around it in smaller range, and denoising effect is also typically directly proportional to participating in the similar pixel quantity of denoising, therefore draw
Enter non local way of search to be significant to lifting denoising performance.The accuracy of similarity is that non-Local Search plays a role
Basic guarantee, if the accuracy of similarity can not be guaranteed, using image block vector between Gauss weighting it is European away from as
Similarity between pixel, has been obviously improved the robustness of similarity in noise circumstance.
Noise in noisy image is typically considered additive white noise of the average for the poor position of zero standard, therefore, rationally
Ground estimation noise level can play certain guiding function for denoising, and the noise level of smooth region often compares in image
The abundant region of details is closer to the noise level of image, therefore related homologous of the picture material that is obtained using super-pixel segmentation
Block, it is to avoid the influence at edge, meanwhile, combining information entropy evaluates the smoothness of super-pixel homology region, and information entropy gets over
Texture information in small representative image is fewer, then image is more smooth.Therefore, more smooth super-pixel homology region is selected to be used for
Estimate the noise level of image, NLM Denoising Algorithms are corrected on this basis, denoising more effectively is carried out to image, to image
Carry out remaining details and marginal information well while denoising.
The content of the invention
It is a kind of adaptive it is an object of the invention to propose, it is easy to accomplish, the image de-noising method of strong robustness, so that
Conventional images denoising method is solved the problem of denoising degree deficiency caused by unknown noise level or excessive denoising, and
Edge and texture information are effectively remained while denoising.
The purpose of the present invention is achieved through the following technical solutions：A kind of method of image denoising, this method include with
Lower step：
(1) the figure J that made an uproar to band carries out the super-pixel segmentation based on entropy rate, obtains some super-pixel homology region Li；
(2) according to image information entropy theory, the comentropy of each super-pixel homology region is calculated：
LiI-th of super-pixel homology region is represented, l represents the gray level of pixel,Represent i-th of super-pixel homologous region
Gray level is l probability in domain,Represent LiIn comentropy；
(3) size of the comentropy of each super-pixel homology region obtained according to calculating is ranked up, and selects value minimum
T super-pixel homology region the 10~20% of sum (account for), subregion the most smooth as in image, and labeled as 1,
Other zone markers are 0, obtain the binary map of smooth region；
(4) standard deviation of each smooth homology region is calculated
It is LiMiddle pixel x intensity,Represent LiMiddle mean intensity,For the number of pixel in super-pixel block；
(5) average for calculating above-mentioned smooth homology region is poor, the noise of the entire image obtained in this, as estimation
Level：
(6) the noise level amendment NLM Denoising Algorithms obtained using estimation carry out denoising to noisy image, obtain denoising figure
F, wherein, the intensity F (x) of the pixel x in denoising image F is：
I represents band and made an uproar the neighborhood of pixel x in figure J, and y represents the pixel in neighborhood I, and J (y) represents pixel y in band makes an uproar figure J
Intensity；W (x, y) represents the similitude between pixel x and pixel y：
Z (x) is the normalized parameter at pixel x, and h represents the smoothing parameter of control denoising degree,
C is constant, C ∈ [300,500] (16).
The beneficial effects of the present invention are：The same source block of the picture material correlation obtained using super-pixel segmentation, it is to avoid
The influence at edge, meanwhile, combining information entropy evaluates the smoothness of super-pixel homology region, the smaller representative image of information entropy
In texture information it is fewer, then image is more smooth.Therefore, more smooth super-pixel homology region is selected to be used for estimating image
Noise level, corrects NLM Denoising Algorithms on this basis, more effectively carries out denoising to image, and denoising is being carried out to image
Details and marginal information are remained well simultaneously.
Brief description of the drawings
Fig. 1 is the FB(flow block) of inventive method.
Fig. 2 is any noisy image.
Fig. 3 is that noisy image carries out the segmentation figure that super-pixel segmentation is obtained.
Fig. 4 is by calculating the binary map that the comentropy of homology region is obtained.
Fig. 5 is Fig. 2 final denoising effect figure.
Fig. 6~9 are that band is made an uproar figure, three ratios using NLM method denoising figures and the inventive method (S+NLM) denoising figure
Compared with exemplary plot.
Embodiment
The present invention it is a kind of based on noise estimate image de-noising method, first with super-pixel segmentation according to picture material by its
It is divided into some homology regions, finds out homology region more smooth in image by foundation of image information entropy, and estimate smooth area
The noise criteria in domain is poor, so as to be modified according to noise level to non-local mean (NLM) denoising method, realizes according to noise
Degree reasonably controls denoising degree, and details is preferably remained while denoising.
It is described in detail below in conjunction with the accompanying drawings with example：
Fig. 1 is the simple process block diagram of the inventive method.With reference to embodiment, the invention will be further described.
(1) obtain a band to make an uproar figure J, as shown in Figure 2；
(2) super-pixel segmentation based on entropy rate is carried out to image, obtained in some super-pixel homology regions, the present embodiment,
200 homology regions are partitioned into altogether, as shown in Figure 3；
(3) according to image information entropy theory, comentropy of each super-pixel with source block is calculated：
LiI-th of super-pixel homology region is represented, l represents the gray level of pixel,Represent i-th of super-pixel homologous region
Domain LiMiddle gray level is l probability,Represent LiComentropy；
(4) size of the comentropy of each super-pixel homology region obtained according to calculating is ranked up, and selects value minimum
20 super-pixel homology regions, be to select 10 in subregion the most smooth, the present embodiment as in image
Example, and be 1 by this homeologous zone marker, other zone markers are 0, obtain the binary map of smooth region, as shown in Figure 4；
(5) standard deviation of each smooth homology region is calculated：
It is LiPixel x intensity in super-pixel block,Represent LiMean intensity in super-pixel block,For super-pixel
The number of pixel in block；
(6) average for calculating above-mentioned smooth homology region is poor, the noise of the entire image obtained in this, as estimation
Level：
It is 20 that t, which is represented in the number of selected super-pixel homology region, the present embodiment,；
(7) the noise level amendment NLM Denoising Algorithms obtained using estimation carry out denoising to noisy image, obtain Fig. 5 institutes
The final denoising figure shown.Calculate intensity of the pixel x in denoising image F：
I represents band and made an uproar the neighborhood of pixel x in figure J, and y represents the pixel in neighborhood I, and J (y) represents pixel y in band makes an uproar figure J
Intensity；W (x, y) represents the similitude between pixel x and pixel y, and its value is decided by the rectangular area centered on x and y
Distance, calculation formula is as follows：
Z (x) is normalized parameter, and h represents with controlling denoising degree smoothing parameter, here according to estimated noise water
Put down to correct smoothing parameter,
C is that value is 300 in constant, the present embodiment, and denoising degree is controlled by the noise level for estimating to obtain, with
To more preferable denoising effect.
A kind of denoising method estimated based on noise of the present invention, is classified as first with super-pixel segmentation according to picture material
Some homology regions, homology region more smooth in image are found out by foundation of image information entropy, and estimate more smooth area
The noise criteria in domain is poor, using the noise criteria difference of these smooth regions as the noise level of entire image, so that according to noise
Level is modified to non-local mean (NLM) denoising method, and realization reasonably controls denoising degree according to noise level, can
Adaptively handle noisy image.The denoising method of denoising and the present invention are carried out to such as Fig. 6 according to existing NLM methods respectively
4 width noisy images shown in~9 carry out denoising, it can be seen that the image whole structure after denoising is more traditional
NLM algorithms are significantly lifted, and details is preferably remained while denoising.
Claims (1)
1. a kind of image de-noising method estimated based on noise, it is characterised in that this method comprises the following steps：
(1) the figure J that made an uproar to band carries out the super-pixel segmentation based on entropy rate, obtains some super-pixel homology region Li；
(2) according to image information entropy theory, the comentropy of each super-pixel homology region is calculated：
<mrow>
<msub>
<mi>H</mi>
<msub>
<mi>L</mi>
<mi>i</mi>
</msub>
</msub>
<mo>=</mo>
<mo>-</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>l</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mn>255</mn>
</munderover>
<msub>
<mi>p</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>L</mi>
<mi>i</mi>
</msub>
<mo>,</mo>
<mi>l</mi>
<mo>)</mo>
</mrow>
</msub>
<mi>l</mi>
<mi>n</mi>
<mi> </mi>
<msub>
<mi>p</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>L</mi>
<mi>i</mi>
</msub>
<mo>,</mo>
<mi>l</mi>
<mo>)</mo>
</mrow>
</msub>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
</mrow>
LiI-th of super-pixel homology region is represented, l represents the gray level of pixel,Represent in i-th of super-pixel homology region
Gray level is l probability,Represent LiIn comentropy；
(3) size of the comentropy of each super-pixel homology region obtained according to calculating is ranked up, and selects value minimum t
Super-pixel homology region the 10~20% of sum (account for), subregion the most smooth as in image, and labeled as 1, other
Zone marker is 0, obtains the binary map of smooth region；
(4) standard deviation of each smooth homology region is calculated
<mrow>
<msub>
<mi>&sigma;</mi>
<msub>
<mi>L</mi>
<mi>i</mi>
</msub>
</msub>
<mo>=</mo>
<msqrt>
<mrow>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>x</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<msub>
<mi>n</mi>
<msub>
<mi>L</mi>
<mi>i</mi>
</msub>
</msub>
</munderover>
<msup>
<mrow>
<mo>(</mo>
<msub>
<mi>I</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>L</mi>
<mi>i</mi>
</msub>
<mo>,</mo>
<mi>x</mi>
<mo>)</mo>
</mrow>
</msub>
<mo>-</mo>
<msub>
<mi>&mu;</mi>
<msub>
<mi>L</mi>
<mi>i</mi>
</msub>
</msub>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
<msub>
<mi>n</mi>
<msub>
<mi>L</mi>
<mi>i</mi>
</msub>
</msub>
</mrow>
</msqrt>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>2</mn>
<mo>)</mo>
</mrow>
</mrow>
<mrow>
<msub>
<mi>&mu;</mi>
<msub>
<mi>L</mi>
<mi>i</mi>
</msub>
</msub>
<mo>=</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>x</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<msub>
<mi>n</mi>
<msub>
<mi>L</mi>
<mi>i</mi>
</msub>
</msub>
</munderover>
<msub>
<mi>I</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>L</mi>
<mi>i</mi>
</msub>
<mo>,</mo>
<mi>x</mi>
<mo>)</mo>
</mrow>
</msub>
<mo>/</mo>
<msub>
<mi>n</mi>
<msub>
<mi>L</mi>
<mi>i</mi>
</msub>
</msub>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>3</mn>
<mo>)</mo>
</mrow>
</mrow>
It is LiMiddle pixel x intensity,Represent LiMiddle mean intensity,For the number of pixel in super-pixel block.
(5) average for calculating above-mentioned smooth homology region is poor, the noise level of the entire image obtained in this, as estimation：
<mrow>
<mover>
<mi>&sigma;</mi>
<mo>&OverBar;</mo>
</mover>
<mo>=</mo>
<munderover>
<mo>&Sigma;</mo>
<mn>1</mn>
<mi>t</mi>
</munderover>
<msub>
<mi>n</mi>
<msub>
<mi>L</mi>
<mi>i</mi>
</msub>
</msub>
<mo>&times;</mo>
<msub>
<mi>&sigma;</mi>
<msub>
<mi>L</mi>
<mi>i</mi>
</msub>
</msub>
<mo>/</mo>
<munderover>
<mo>&Sigma;</mo>
<mn>1</mn>
<mi>t</mi>
</munderover>
<msub>
<mi>n</mi>
<msub>
<mi>L</mi>
<mi>i</mi>
</msub>
</msub>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>4</mn>
<mo>)</mo>
</mrow>
</mrow>
(6) the noise level amendment NLM Denoising Algorithms obtained using estimation carry out denoising to noisy image, obtain denoising figure F, its
In, the intensity F (x) of the pixel x in denoising image F is：
<mrow>
<mi>F</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<munder>
<mo>&Sigma;</mo>
<mrow>
<mi>y</mi>
<mo>&Element;</mo>
<mi>I</mi>
</mrow>
</munder>
<mi>w</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mi>J</mi>
<mrow>
<mo>(</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>5</mn>
<mo>)</mo>
</mrow>
</mrow>
I represents band and made an uproar the neighborhood of pixel x in figure J, and y represents the pixel in neighborhood I, and J (y) represents that pixel y is strong in band makes an uproar figure J
Degree；W (x, y) represents the similitude between pixel x and pixel y.
<mrow>
<mi>w</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfrac>
<mn>1</mn>
<mrow>
<mi>Z</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>)</mo>
</mrow>
</mrow>
</mfrac>
<mi>exp</mi>
<mrow>
<mo>(</mo>
<mo>-</mo>
<mfrac>
<mrow>
<mo>|</mo>
<mo>|</mo>
<mi>V</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>)</mo>
</mrow>
<mo>-</mo>
<mi>V</mi>
<mrow>
<mo>(</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>|</mo>
<msup>
<mo>|</mo>
<mn>2</mn>
</msup>
</mrow>
<msup>
<mi>h</mi>
<mn>2</mn>
</msup>
</mfrac>
<mo>)</mo>
</mrow>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>6</mn>
<mo>)</mo>
</mrow>
</mrow>
<mrow>
<mi>Z</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<munder>
<mo>&Sigma;</mo>
<mrow>
<mi>y</mi>
<mo>&Element;</mo>
<mi>I</mi>
</mrow>
</munder>
<mi>exp</mi>
<mrow>
<mo>(</mo>
<mo>-</mo>
<mfrac>
<mrow>
<mo>|</mo>
<mo>|</mo>
<mi>V</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>)</mo>
</mrow>
<mo>-</mo>
<mi>V</mi>
<mrow>
<mo>(</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>|</mo>
<msup>
<mo>|</mo>
<mn>2</mn>
</msup>
</mrow>
<msup>
<mi>h</mi>
<mn>2</mn>
</msup>
</mfrac>
<mo>)</mo>
</mrow>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>7</mn>
<mo>)</mo>
</mrow>
</mrow>
Z (x) is the normalized parameter at pixel x, and h represents the smoothing parameter of control denoising degree,
C is constant, C ∈ [300,500] (8).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710390337.1A CN107330863B (en) | 2017-05-27 | 2017-05-27 | A kind of image de-noising method based on noise estimation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710390337.1A CN107330863B (en) | 2017-05-27 | 2017-05-27 | A kind of image de-noising method based on noise estimation |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107330863A true CN107330863A (en) | 2017-11-07 |
| CN107330863B CN107330863B (en) | 2019-10-18 |
Family
ID=60192961
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710390337.1A Expired - Fee Related CN107330863B (en) | 2017-05-27 | 2017-05-27 | A kind of image de-noising method based on noise estimation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107330863B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109064418A (en) * | 2018-07-11 | 2018-12-21 | 成都信息工程大学 | A kind of Images Corrupted by Non-uniform Noise denoising method based on non-local mean |
| CN109544468A (en) * | 2018-10-26 | 2019-03-29 | 浙江师范大学 | A kind of image data amplification method |
| WO2022246663A1 (en) * | 2021-05-25 | 2022-12-01 | 深圳市大疆创新科技有限公司 | Image processing method, device and system, and storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101661611A (en) * | 2009-09-25 | 2010-03-03 | 西安电子科技大学 | Realization method based on bayesian non-local mean filter |
| CN102298773B (en) * | 2011-09-19 | 2012-12-26 | 西安电子科技大学 | Shape-adaptive non-local mean denoising method |
| CN103116873A (en) * | 2013-03-20 | 2013-05-22 | 中国科学院深圳先进技术研究院 | Image noise reducing method |
| CN103208105A (en) * | 2013-05-02 | 2013-07-17 | 中国电子科技集团公司第二十八研究所 | Infrared image detail enhancement and noise adaptive inhibition method |
| US20150138405A1 (en) * | 2013-11-15 | 2015-05-21 | Linear Algebra Technologies Limited | Apparatus, systems, and methods for removing noise from an image |
| CN105913396A (en) * | 2016-04-11 | 2016-08-31 | 湖南源信光电科技有限公司 | Noise estimation-based image edge preservation mixed de-noising method |
-
2017
- 2017-05-27 CN CN201710390337.1A patent/CN107330863B/en not_active Expired - Fee Related
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101661611A (en) * | 2009-09-25 | 2010-03-03 | 西安电子科技大学 | Realization method based on bayesian non-local mean filter |
| CN102298773B (en) * | 2011-09-19 | 2012-12-26 | 西安电子科技大学 | Shape-adaptive non-local mean denoising method |
| CN103116873A (en) * | 2013-03-20 | 2013-05-22 | 中国科学院深圳先进技术研究院 | Image noise reducing method |
| CN103208105A (en) * | 2013-05-02 | 2013-07-17 | 中国电子科技集团公司第二十八研究所 | Infrared image detail enhancement and noise adaptive inhibition method |
| US20150138405A1 (en) * | 2013-11-15 | 2015-05-21 | Linear Algebra Technologies Limited | Apparatus, systems, and methods for removing noise from an image |
| CN105913396A (en) * | 2016-04-11 | 2016-08-31 | 湖南源信光电科技有限公司 | Noise estimation-based image edge preservation mixed de-noising method |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109064418A (en) * | 2018-07-11 | 2018-12-21 | 成都信息工程大学 | A kind of Images Corrupted by Non-uniform Noise denoising method based on non-local mean |
| CN109064418B (en) * | 2018-07-11 | 2022-03-08 | 成都信息工程大学 | Non-local mean value-based non-uniform noise image denoising method |
| CN109544468A (en) * | 2018-10-26 | 2019-03-29 | 浙江师范大学 | A kind of image data amplification method |
| CN109544468B (en) * | 2018-10-26 | 2023-10-13 | 浙江师范大学 | Image data amplification method |
| WO2022246663A1 (en) * | 2021-05-25 | 2022-12-01 | 深圳市大疆创新科技有限公司 | Image processing method, device and system, and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107330863B (en) | 2019-10-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US8131098B2 (en) | Image processing device, image processing method, image processing system, program, storage medium, and integrated circuit | |
| Kim et al. | Optimized contrast enhancement for real-time image and video dehazing | |
| CN105654438A (en) | Gray scale image fitting enhancement method based on local histogram equalization | |
| Wang et al. | Variational single nighttime image haze removal with a gray haze-line prior | |
| CN109816608B (en) | Low-illumination image self-adaptive brightness enhancement method based on noise suppression | |
| CN104036465A (en) | Edge detection based on self-adaptive nonlocal mean denoising method | |
| CN107330863A (en) | A kind of image de-noising method estimated based on noise | |
| Hou et al. | Underwater image dehazing and denoising via curvature variation regularization | |
| CN102208101A (en) | Self-adaptive linearity transformation enhancing method of infrared image | |
| CN102306307B (en) | Positioning method of fixed point noise in color microscopic image sequence | |
| CN109410235B (en) | Target tracking method fusing edge features | |
| CN105741245A (en) | Grayscale conversion based adaptive contrast enhancement algorithm | |
| CN109754440A (en) | A kind of shadow region detection method based on full convolutional network and average drifting | |
| CN111476744B (en) | Underwater image enhancement method based on classification and atmospheric imaging model | |
| CN110111342B (en) | Optimized selection method and device for matting algorithm | |
| CN109978858B (en) | Double-frame thumbnail image quality evaluation method based on foreground detection | |
| CN109903275B (en) | Fermented grain mildewing area detection method based on self-adaptive multi-scale filtering and histogram comparison | |
| CN117541582B (en) | IGBT insulation quality detection method for high-frequency converter | |
| CN108550124B (en) | Illumination compensation and image enhancement method based on bionic spiral | |
| WO2020107308A1 (en) | Low-light-level image rapid enhancement method and apparatus based on retinex | |
| CN110246169B (en) | Gradient-based window adaptive stereo matching method and system | |
| JP5327766B2 (en) | Memory color correction in digital images | |
| CN111640068A (en) | Unsupervised automatic correction method for image exposure | |
| CN113012079B (en) | Low-brightness vehicle bottom image enhancement method and device and storage medium | |
| CN110796609B (en) | Low-light image enhancement method based on scale perception and detail enhancement model |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20191018 Termination date: 20200527 |