CN109978774A - Multiframe continuously waits the denoising fusion method and device of exposure images - Google Patents
Multiframe continuously waits the denoising fusion method and device of exposure images Download PDFInfo
- Publication number
- CN109978774A CN109978774A CN201711449537.6A CN201711449537A CN109978774A CN 109978774 A CN109978774 A CN 109978774A CN 201711449537 A CN201711449537 A CN 201711449537A CN 109978774 A CN109978774 A CN 109978774A
- Authority
- CN
- China
- Prior art keywords
- image
- block
- corresponding blocks
- noise
- images
- 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
- 238000007500 overflow downdraw method Methods 0.000 title claims abstract description 11
- 238000000034 method Methods 0.000 claims abstract description 31
- 230000004927 fusion Effects 0.000 claims abstract description 22
- 238000010276 construction Methods 0.000 claims description 11
- 238000004364 calculation method Methods 0.000 abstract description 12
- 238000007499 fusion processing Methods 0.000 abstract description 6
- 238000010586 diagram Methods 0.000 description 6
- 238000001914 filtration Methods 0.000 description 5
- 230000006641 stabilisation Effects 0.000 description 3
- 238000011105 stabilization Methods 0.000 description 3
- 239000000654 additive Substances 0.000 description 1
- 230000000996 additive effect Effects 0.000 description 1
- 230000002146 bilateral effect Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 239000002131 composite material Substances 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000003706 image smoothing Methods 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000003786 synthesis reaction Methods 0.000 description 1
- 230000009466 transformation Effects 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
- G06T5/50—Image enhancement or restoration by the use of more than one image, e.g. averaging, subtraction
-
- G06T5/70—
-
- 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
- G06T2207/10—Image acquisition modality
- G06T2207/10016—Video; Image sequence
-
- 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
- G06T2207/20212—Image combination
- G06T2207/20221—Image fusion; Image merging
Abstract
The present invention provides the denoising fusion method and device of a kind of continuous equal exposure images of multiframe, it is intended to which multiple image informations for combining Same Scene carry out noise reduction and synthesize a low noise high quality graphic.The described method includes: choosing a frame image as reference picture from multiframe continuously equal exposure images；The reference picture is subjected to piecemeal, multiple reference blocks is obtained, finds corresponding blocks corresponding with each reference block in other each frame images, obtain multiple series of images block；Every group of image block is merged respectively, obtains the fused image block of multiple groups, and finally obtain the image after Time Domain Fusion.Calculating of the present invention in fusion process greatly reduces calculation amount, has combined computational efficiency and fusion results based entirely on the processing of non-frequency domain.
Description
Technical field
The present invention relates to the denoising fusion sides that technical field of image processing more particularly to a kind of multiframe continuously wait exposure images
Method and device.
Background technique
It is limited to the size of the lens of camera and pixel on mobile phone, under low-light environment, mobile phone photograph is easier
Generate the serious image of noise.How fast and effeciently to estimate noise, removal noise, can also stablize imaging just in low bright situation
It is a challenge.
Mean filter based on point, which can effectively eliminate to be added on each point of image smoothing region, obeys the random noise with distribution,
Similarly, the joint denoising based on several similar image blocks also can highly desirable Removing Random No.Currently, similar based on block
The joint denoising method of property mainly includes 3DNR (3D noise reduction), NLM (non-local mean), BM3D (three-dimensional Block- matching) etc., these
Method is proved to be a kind of highly effective means that picture quality can be improved.And the denoising of image, it is generally divided into two kinds
Implementation: one is in spatial domain directly to the processing of pixel；It is another then first image is converted, then in transform domain
(frequency domain) handles the transformation coefficient of image, it is last it is inverse transformed handled after image.The 3DNR that generallys use with
NLM implementation method is to complete in spatial domain, and BM3D takes into account time-domain and frequency-domain processing.
A kind of protocol procedures used in the prior art are similar to BM3D, main to realize that process includes grouping (alignment) and filters
Two step of wave (merging).Detailed process are as follows: continuous low exposure snap obtains one group of Bayer (Bayer) domain image as input, chooses
Reference frame (lucky image), the judgement after piecemeal again based on distance between pixels in block are each fritter of reference frame image
Corresponding fritter is found in other frames, using the filtering for carrying out time domain and airspace to each group of reference block and corresponding blocks, is obtained
Obtain final output image.The advantage of the program is: in half-light or has under the scene of fast moving objects, system is stablized, at
Picture noise is small, and artifact problem can be controlled very well；And the operation in the domain Bayer can effectively improve the dynamic model of image
It encloses.
But no matter completed in frequency domain in time domain or spatial processing since the committed step in program process merges
, thus have the shortcomings that some protrusions and obvious, mainly have: 1, time-consuming: four planes of the processing of the domain Bayer input picture point
The filtering processing of frequency domain is carried out respectively, and calculation amount becomes larger；2, the standalone processes of frequency domain will affect the value of entire spatial domain pixel,
Its Parameter Modulation and control are more difficult, are easy to generate pseudo-colours；In high contrast and high-frequency region, easily there are various artifacts.
Summary of the invention
Multiframe provided by the invention continuously waits the denoising fusion method and device of exposure images, having been calculated in fusion process
Processing entirely based on non-frequency domain, greatly reduces calculation amount, has combined computational efficiency and fusion results.
In a first aspect, the present invention provides a kind of denoising fusion method of the continuous equal exposure images of multiframe, comprising:
A frame image is chosen as reference picture from multiframe continuously equal exposure images；
The reference picture is subjected to piecemeal, multiple reference blocks is obtained, is found and each reference in other each frame images
The corresponding corresponding blocks of block, obtain multiple series of images block；
Every group of image block is merged respectively, obtains the fused image block of multiple groups, and finally obtain through Time Domain Fusion
Image afterwards.
Optionally, every group of image block is merged respectively described, obtains the fused image block of multiple groups, and final
To after the image after Time Domain Fusion, the method also includes:
Airspace filter is carried out to the image after Time Domain Fusion.
Optionally, described to merge to every group of image block, obtaining the fused image block of multiple groups includes:
Calculate the distance between the corresponding blocks in the reference picture in each reference block and other each frame images；
The distance calculated is compared with preset threshold；
When the distance calculated is less than preset threshold, keep the corresponding blocks in other frame images constant；
When the distance calculated is greater than preset threshold, the automatic corresponding blocks for constructing Noise；
It is weighted and averaged all corresponding blocks and reference block, obtains fused image block.
Optionally, the corresponding blocks of the automatic construction Noise include:
The number m of statistics and the unmatched corresponding blocks of reference block；
Obtain the noise level estimation of reference block image；
M corresponding blocks of automatic construction Noise.
Optionally, m corresponding blocks of the automatic construction Noise include:
Estimate original noiseless reference block；
Random noise is added for the original noiseless reference block, constructs corresponding blocks m of Noise.
Second aspect, the present invention provide a kind of denoising fusing device of the continuous equal exposure images of multiframe, comprising:
Selection unit, for choosing a frame image as reference picture from multiframe continuously equal exposure images；
Grouped element obtains multiple reference blocks, looks in other each frame images for the reference picture to be carried out piecemeal
To corresponding blocks corresponding with each reference block, multiple series of images block is obtained；
Integrated unit obtains the fused image block of multiple groups for merging respectively to every group of image block, and final
To the image after Time Domain Fusion
Optionally, described device further include:
Filter unit, for carrying out airspace filter to the image after Time Domain Fusion.
Optionally, the integrated unit includes:
Computing module, for calculating between the corresponding blocks in the reference picture in each reference block and other each frame images
Distance；
Comparison module, for the distance calculated to be compared with preset threshold；
Module is kept, for keeping the corresponding blocks in other frame images when the distance calculated is less than preset threshold
It is constant；
Constructing module, for constructing the corresponding blocks of Noise automatically when the distance calculated is greater than preset threshold；
Weighting block obtains fused image block for being weighted and averaged all corresponding blocks and reference block.
Optionally, the constructing module includes:
Statistic submodule, for counting and the number m of the unmatched corresponding blocks of reference block；
Acquisition submodule, the noise level for obtaining reference block image are estimated；
Submodule is constructed, for constructing m corresponding blocks of Noise automatically.
Optionally, the construction submodule includes:
Sub-module is estimated, for estimating original noiseless reference block；
Sub-module is constructed, for adding random noise for the original noiseless reference block, constructs the correspondence of Noise
Block m.
Multiframe provided in an embodiment of the present invention continuously waits the denoising fusion method and device of exposure images, to avoid at frequency domain
Reason completes merging in time domain and airspace, the meter in fusion process to input picture directly to image group pixel operation in two steps
The processing based entirely on non-frequency domain is calculated, calculation amount is greatly reduced, both can calculate speed to avoid the time-consuming of frequency-domain calculations to improve
Degree, has combined computational efficiency and fusion results；Also the various artifacts introduced because frequency domain is dealt with improperly are avoided that, thus
Easy to control/system stabilization etc. is improved.
Detailed description of the invention
Fig. 1 is the flow chart for the denoising fusion method that one embodiment of the invention multiframe continuously waits exposure images；
Fig. 2 is the structural schematic diagram for the denoising fusing device that one embodiment of the invention multiframe continuously waits exposure images；
Fig. 3 is the structural schematic diagram for the denoising fusing device that another embodiment of the present invention multiframe continuously waits exposure images；
Fig. 4 is the structural schematic diagram for the denoising fusing device that yet another embodiment of the invention multiframe continuously waits exposure images；
Fig. 5 is the structural schematic diagram for the denoising fusing device that further embodiment of this invention multiframe continuously waits exposure images；
Fig. 6 is the structural schematic diagram for the denoising fusing device that a further embodiment multiframe of the present invention continuously waits exposure images.
Specific embodiment
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with the embodiment of the present invention
In attached drawing, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described embodiment is only
It is only a part of the embodiment of the present invention, instead of all the embodiments.Based on the embodiments of the present invention, ordinary skill
Personnel's every other embodiment obtained without making creative work, shall fall within the protection scope of the present invention.
Multiple image informations for being intended to joint Same Scene carry out noise reduction and synthesize a low noise high quality graphic, the present invention
Embodiment provides a kind of denoising fusion method of the continuous equal exposure images of multiframe, as shown in Figure 1, which comprises
S11, a frame image is chosen as reference picture from the exposure images such as multiframe is continuous.
S12, by the reference picture carry out piecemeal, obtain multiple reference blocks, found in other each frame images with each
The corresponding corresponding blocks of reference block, obtain multiple series of images block.
S13, every group of image block is merged respectively, obtains the fused image block of multiple groups, and finally obtain through time domain
Fused image.
Multiframe provided in an embodiment of the present invention continuously waits the denoising fusion method of exposure images, to avoid frequency domain from handling, directly
It connects to image group pixel operation, completes merging in time domain and airspace, having been calculated in fusion process to input picture in two steps
Processing entirely based on non-frequency domain, greatly reduces calculation amount, both can to avoid the time-consuming of frequency-domain calculations, to improve calculating speed,
Computational efficiency and fusion results are combined；Also the various artifacts introduced because frequency domain is dealt with improperly are avoided that, thus easy
Control/system stabilization etc. is improved.
Continuously the denoising fusion method of exposure images is waited to be described in detail multiframe of the present invention below.
One group of S21, input image.
Here image refers to one group of image continuous and with identical shorter exposure time, is suitable for noise model letter
The image (such as domain Bayer) that Dan Yi estimates.
S22, reference picture is chosen, and realizes that piecemeal is aligned.
An image is chosen from multiple image as reference picture, and reference picture is then subjected to piecemeal, is based on Gauss
Top and bottom process is found in each frame correspondence image from slightly to thin for each small images (tile, reference block) in reference picture
Corresponding fritter (corresponding blocks), this is the process of a grouping in fact.
S23, multiframe figure segment fusion.
Based on step S22 grouping as a result, merging each group of image fritter, specific step is as follows:
The distance between fritter is corresponded in each fritter and other images in S231, calculating reference picture, can refer to public affairs
Formula (1):
Wherein, m*n is the size of image fritter, TcWith Ta,lRefer respectively to the corresponding blocks in reference block and l frame, distance D
It is not limited to 2- norm, is also possible to 1- norm, the i.e. sum of the absolute value of difference or other definition norms.
S232, the size for comparing D Yu preset threshold T.
D is the calculated result of formula (1), and threshold value T is a default value.Comparison result is handled in two kinds of situation:
(1) D≤T: illustrate that corresponding blocks are similar to reference block content, go to step S233；
(2) D > T: illustrating that corresponding blocks content in the i-th frame and reference block difference are larger, and the corresponding blocks are directly given up in selection,
Then step S234 is gone to.
S233, keep the corresponding blocks in the frame constant.
The corresponding blocks of S234, automatic construction Noise.
The reasons why taking the step is, rejects inappropriate piece, can prevent occurring artifact in fusion process；It constructs noisy
The corresponding blocks of sound are then to eliminate the noise level non-uniform phenomenon of composograph.The detailed process of the step includes:
A, the number m of statistics and the unmatched corresponding blocks of reference block
B, the noise level estimation of reference block image, noise level (variance) σ are obtained1 2Estimation can be by asking reference block
Mean square deviation and obtain (based on image noise model simply with fritter divide premise)
C, the m of construction Noise opens corresponding blocks automatically:
C1, the original noiseless reference block T of estimations, the mode that can be taken is: 1, being completed based on Gauss nuclear convolution to reference block
Filtering；2, average using other corresponding blocks less than threshold value and reference block, the image after obtaining noise reduction.Method should be not limited to this
It two kinds, can also be in conjunction with image after the image information acquisition noise reduction of interframe and frame between interior.
C2, formula (2) are based on, give TsUpper addition random noise, the automatic corresponding blocks for constructing Noise m.
Ta,i=Ts+σ2*randn(row,col) (2)
The domain Bayer noise at dark field based on additive white noise, and brightness increase when be mainly shown as Poisson distribution
Model, for ease of calculation, in conjunction with the characteristics of image fritter, it is assumed that the noise level in each small images is identical, noise
Variance is to estimate to obtain based on step b, then obtains final variance, i.e. σ by appropriateness scaling again2=k* σ1 2, wherein k be
Adjustment factor, randn (row, col) are responsible for generating random Gaussian.
S235, all corresponding blocks of weighted average and reference block.
Formula (3) gives simplest direct averagely adduction calculating method.But calculation method is without being limited thereto, can also use
Other methods, such as the weight computing method based on distance
W in formula (4)iAs weight；The weighting method based on node-by-node algorithm can also be used, as shown in formula (5):
Realize that image fritter merges by group based on above step, the final fusion for realizing entire image.
S24, it exports composite diagram and completes airspace denoising.
For the operation for all carrying out step S23 with reference to image fritter all on figure, figure of the width after Time Domain Fusion is obtained
Picture, the step mainly complete the denoising of multiple images in the time domain with merge, for further denoising, select NLM method or
Bilateral filtering method or other methods complete the filtering in airspace.
Image after S25, output synthesis.
The embodiment of the present invention also provides a kind of denoising fusing device of the continuous equal exposure images of multiframe, as shown in Fig. 2, described
Device includes:
Selection unit 11, for choosing a frame image as reference picture from multiframe continuously equal exposure images；
Grouped element 12 obtains multiple reference blocks, in other each frame images for the reference picture to be carried out piecemeal
Corresponding blocks corresponding with each reference block are found, multiple series of images block is obtained；
Integrated unit 13 obtains the fused image block of multiple groups, and final for merging respectively to every group of image block
Obtain the image after Time Domain Fusion
Multiframe provided in an embodiment of the present invention continuously waits the denoising fusing device of exposure images, to avoid frequency domain from handling, directly
It connects to image group pixel operation, completes merging in time domain and airspace, having been calculated in fusion process to input picture in two steps
Processing entirely based on non-frequency domain, greatly reduces calculation amount, both can to avoid the time-consuming of frequency-domain calculations, to improve calculating speed,
Computational efficiency and fusion results are combined；Also the various artifacts introduced because frequency domain is dealt with improperly are avoided that, thus easy
Control/system stabilization etc. is improved.
Further, as shown in figure 3, described device further include:
Filter unit 14, for carrying out airspace filter to the image after Time Domain Fusion.
Optionally, as shown in figure 4, the integrated unit 13 includes:
Computing module 131, for calculating the corresponding blocks in the reference picture in each reference block and other each frame images
The distance between；
Comparison module 132, for the distance calculated to be compared with preset threshold；
Module 133 is kept, for keeping the correspondence in other frame images when the distance calculated is less than preset threshold
Block is constant；
Constructing module 134, for constructing the correspondence of Noise automatically when the distance calculated is greater than preset threshold
Block；
Weighting block 135 obtains fused image block for being weighted and averaged all corresponding blocks and reference block.
Optionally, as shown in figure 5, the constructing module 134 includes:
Statistic submodule 1341, for counting and the number m of the unmatched corresponding blocks of reference block；
Acquisition submodule 1342, the noise level for obtaining reference block image are estimated；
Submodule 1343 is constructed, for constructing m corresponding blocks of Noise automatically.
Optionally, as shown in fig. 6, the construction submodule 1343 includes:
Sub-module 13431 is estimated, for estimating original noiseless reference block；
Sub-module 13432 is constructed, for adding random noise for the original noiseless reference block, constructs Noise
Corresponding blocks m.
The device of the present embodiment can be used for executing the technical solution of above method embodiment, realization principle and technology
Effect is similar, and details are not described herein again.
Those of ordinary skill in the art will appreciate that realizing all or part of the process in above-described embodiment method, being can be with
Relevant hardware is instructed to complete by computer program, the program can be stored in a computer-readable storage medium
In, the program is when being executed, it may include such as the process of the embodiment of above-mentioned each method.Wherein, the storage medium can be magnetic
Dish, CD, read-only memory (Read-Only Memory, ROM) or random access memory (Random Access
Memory, RAM) etc..
The above description is merely a specific embodiment, but scope of protection of the present invention is not limited thereto, any
In the technical scope disclosed by the present invention, any changes or substitutions that can be easily thought of by those familiar with the art, all answers
It is included within the scope of the present invention.Therefore, protection scope of the present invention should be subject to the protection scope in claims.
Claims (10)
1. the denoising fusion method that a kind of multiframe continuously waits exposure images characterized by comprising
A frame image is chosen as reference picture from multiframe continuously equal exposure images；
The reference picture is subjected to piecemeal, multiple reference blocks is obtained, is found in other each frame images and each reference block phase
The corresponding blocks answered obtain multiple series of images block；
Every group of image block is merged respectively, obtains the fused image block of multiple groups, and finally obtain after Time Domain Fusion
Image.
2. being obtained the method according to claim 1, wherein being merged respectively to every group of image block described
The fused image block of multiple groups, and after finally obtaining the image after Time Domain Fusion, the method also includes:
Airspace filter is carried out to the image after Time Domain Fusion.
3. method according to claim 1 or 2, which is characterized in that it is described that every group of image block is merged, obtain multiple groups
Fused image block includes:
Calculate the distance between the corresponding blocks in the reference picture in each reference block and other each frame images；
The distance calculated is compared with preset threshold；
When the distance calculated is less than preset threshold, keep the corresponding blocks in other frame images constant；
When the distance calculated is greater than preset threshold, the automatic corresponding blocks for constructing Noise；
It is weighted and averaged all corresponding blocks and reference block, obtains fused image block.
4. according to the method described in claim 3, it is characterized in that, the corresponding blocks of the automatic construction Noise include:
The number m of statistics and the unmatched corresponding blocks of reference block；
Obtain the noise level estimation of reference block image；
M corresponding blocks of automatic construction Noise.
5. according to the method described in claim 4, it is characterized in that, m corresponding blocks of the automatic construction Noise include:
Estimate original noiseless reference block；
Random noise is added for the original noiseless reference block, constructs corresponding blocks m of Noise.
6. the denoising fusing device that a kind of multiframe continuously waits exposure images characterized by comprising
Selection unit, for choosing a frame image as reference picture from multiframe continuously equal exposure images；
Grouped element, for by the reference picture carry out piecemeal, obtain multiple reference blocks, found in other each frame images with
Each corresponding corresponding blocks of reference block, obtain multiple series of images block；
Integrated unit obtains the fused image block of multiple groups for merging respectively to every group of image block, and finally obtain through
Image after Time Domain Fusion.
7. device according to claim 6, which is characterized in that described device further include:
Filter unit, for carrying out airspace filter to the image after Time Domain Fusion.
8. device according to claim 6 or 7, which is characterized in that the integrated unit includes:
Computing module, for calculate between the corresponding blocks in the reference picture in each reference block and other each frame images away from
From；
Comparison module, for the distance calculated to be compared with preset threshold；
Module is kept, for keeping the corresponding blocks in other frame images constant when the distance calculated is less than preset threshold；
Constructing module, for constructing the corresponding blocks of Noise automatically when the distance calculated is greater than preset threshold；
Weighting block obtains fused image block for being weighted and averaged all corresponding blocks and reference block.
9. device according to claim 8, which is characterized in that the constructing module includes:
Statistic submodule, for counting and the number m of the unmatched corresponding blocks of reference block；
Acquisition submodule, the noise level for obtaining reference block image are estimated；
Submodule is constructed, for constructing m corresponding blocks of Noise automatically.
10. device according to claim 9, which is characterized in that the construction submodule includes:
Sub-module is estimated, for estimating original noiseless reference block；
Sub-module is constructed, for adding random noise for the original noiseless reference block, constructs the corresponding blocks m of Noise
It is a.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711449537.6A CN109978774B (en) | 2017-12-27 | 2017-12-27 | Denoising fusion method and device for multi-frame continuous equal exposure images |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711449537.6A CN109978774B (en) | 2017-12-27 | 2017-12-27 | Denoising fusion method and device for multi-frame continuous equal exposure images |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109978774A true CN109978774A (en) | 2019-07-05 |
| CN109978774B CN109978774B (en) | 2021-06-18 |
Family
ID=67071746
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711449537.6A Active CN109978774B (en) | 2017-12-27 | 2017-12-27 | Denoising fusion method and device for multi-frame continuous equal exposure images |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109978774B (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110689502A (en) * | 2019-10-09 | 2020-01-14 | 深圳看到科技有限公司 | Image processing method and related device |
| CN110990088A (en) * | 2019-12-09 | 2020-04-10 | Oppo广东移动通信有限公司 | Data processing method and related equipment |
| CN111986155A (en) * | 2020-07-20 | 2020-11-24 | 广州润普网络科技有限公司 | Universal scene character recognition method and device |
| CN112801908A (en) * | 2021-02-05 | 2021-05-14 | 深圳技术大学 | Image denoising method and device, computer equipment and storage medium |
| CN113191965A (en) * | 2021-04-14 | 2021-07-30 | 浙江大华技术股份有限公司 | Image noise reduction method, device and computer storage medium |
| CN113240607A (en) * | 2021-05-26 | 2021-08-10 | Oppo广东移动通信有限公司 | Image denoising method and device, electronic equipment and storage medium |
| CN113766150A (en) * | 2021-08-31 | 2021-12-07 | 北京安酷智芯科技有限公司 | Noise reduction method, circuit system, electronic equipment and computer readable storage medium |
| WO2023273868A1 (en) * | 2021-06-29 | 2023-01-05 | 展讯通信（上海）有限公司 | Image denoising method and apparatus, terminal, and storage medium |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101639934A (en) * | 2009-09-04 | 2010-02-03 | 西安电子科技大学 | SAR image denoising method based on contour wave domain block hidden Markov model |
| CN102238316A (en) * | 2010-04-29 | 2011-11-09 | 北京科迪讯通科技有限公司 | Self-adaptive real-time denoising scheme for 3D digital video image |
| CN103606132A (en) * | 2013-10-31 | 2014-02-26 | 西安电子科技大学 | Multiframe digital image denoising method based on space domain and time domain combination filtering |
| US20140341434A1 (en) * | 2013-05-17 | 2014-11-20 | Industrial Technology Research Institute | Dymanic fusion method and device of images |
| CN104952040A (en) * | 2014-03-26 | 2015-09-30 | 安凯(广州)微电子技术有限公司 | Image filtering method and image filtering device |
| CN105741255A (en) * | 2014-12-08 | 2016-07-06 | 展讯通信（上海）有限公司 | Image fusion method and device |
| WO2016113888A1 (en) * | 2015-01-15 | 2016-07-21 | 株式会社島津製作所 | Image processing apparatus |
| CN106251365A (en) * | 2016-07-22 | 2016-12-21 | 北京邮电大学 | Many exposure video fusion method and device |
| CN106251318A (en) * | 2016-09-29 | 2016-12-21 | 杭州雄迈集成电路技术有限公司 | A kind of denoising device and method of sequence image |
| US20170091908A1 (en) * | 2014-05-19 | 2017-03-30 | Shimadzu Corporation | Image-processing device |
| CN106897975A (en) * | 2017-02-07 | 2017-06-27 | 信阳师范学院 | A kind of image de-noising method of hypercube Granule Computing |
| CN107403413A (en) * | 2017-04-14 | 2017-11-28 | 杭州当虹科技有限公司 | A kind of video multiframe denoising and Enhancement Method |
-
2017
- 2017-12-27 CN CN201711449537.6A patent/CN109978774B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101639934A (en) * | 2009-09-04 | 2010-02-03 | 西安电子科技大学 | SAR image denoising method based on contour wave domain block hidden Markov model |
| CN102238316A (en) * | 2010-04-29 | 2011-11-09 | 北京科迪讯通科技有限公司 | Self-adaptive real-time denoising scheme for 3D digital video image |
| US20140341434A1 (en) * | 2013-05-17 | 2014-11-20 | Industrial Technology Research Institute | Dymanic fusion method and device of images |
| CN103606132A (en) * | 2013-10-31 | 2014-02-26 | 西安电子科技大学 | Multiframe digital image denoising method based on space domain and time domain combination filtering |
| CN104952040A (en) * | 2014-03-26 | 2015-09-30 | 安凯(广州)微电子技术有限公司 | Image filtering method and image filtering device |
| US20170091908A1 (en) * | 2014-05-19 | 2017-03-30 | Shimadzu Corporation | Image-processing device |
| CN105741255A (en) * | 2014-12-08 | 2016-07-06 | 展讯通信（上海）有限公司 | Image fusion method and device |
| WO2016113888A1 (en) * | 2015-01-15 | 2016-07-21 | 株式会社島津製作所 | Image processing apparatus |
| CN106251365A (en) * | 2016-07-22 | 2016-12-21 | 北京邮电大学 | Many exposure video fusion method and device |
| CN106251318A (en) * | 2016-09-29 | 2016-12-21 | 杭州雄迈集成电路技术有限公司 | A kind of denoising device and method of sequence image |
| CN106897975A (en) * | 2017-02-07 | 2017-06-27 | 信阳师范学院 | A kind of image de-noising method of hypercube Granule Computing |
| CN107403413A (en) * | 2017-04-14 | 2017-11-28 | 杭州当虹科技有限公司 | A kind of video multiframe denoising and Enhancement Method |
Non-Patent Citations (4)
| Title |
|---|
| LEI Z.等: "High quality multi-focus polychromatic composite image fusion algorithm based on filtering in frequency domain and synthesis in space domain", 《JOURNAL OF ZHEJIANG UNIVERSITY SCIENCE C (COMPUT & ELECTRON)》 * |
| MUHAMMAD H.等: "Wavelet based multi-focus image fusion using adaptive sized blocks", 《2009 IEEE 13TH INTERNATIONAL MULTITOPIC CONFERENCE》 * |
| 景华炯 等: "双重自适应非局部均值图像去噪算法", 《南阳理工学院学报》 * |
| 王翔 等: "一种基于多通道联合估计的非局部均值彩色图像去噪方法", 《数据采集与处理》 * |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110689502A (en) * | 2019-10-09 | 2020-01-14 | 深圳看到科技有限公司 | Image processing method and related device |
| CN110990088A (en) * | 2019-12-09 | 2020-04-10 | Oppo广东移动通信有限公司 | Data processing method and related equipment |
| CN110990088B (en) * | 2019-12-09 | 2023-08-11 | Oppo广东移动通信有限公司 | Data processing method and related equipment |
| CN111986155A (en) * | 2020-07-20 | 2020-11-24 | 广州润普网络科技有限公司 | Universal scene character recognition method and device |
| CN112801908A (en) * | 2021-02-05 | 2021-05-14 | 深圳技术大学 | Image denoising method and device, computer equipment and storage medium |
| CN112801908B (en) * | 2021-02-05 | 2022-04-22 | 深圳技术大学 | Image denoising method and device, computer equipment and storage medium |
| CN113191965A (en) * | 2021-04-14 | 2021-07-30 | 浙江大华技术股份有限公司 | Image noise reduction method, device and computer storage medium |
| CN113240607A (en) * | 2021-05-26 | 2021-08-10 | Oppo广东移动通信有限公司 | Image denoising method and device, electronic equipment and storage medium |
| WO2023273868A1 (en) * | 2021-06-29 | 2023-01-05 | 展讯通信（上海）有限公司 | Image denoising method and apparatus, terminal, and storage medium |
| CN113766150A (en) * | 2021-08-31 | 2021-12-07 | 北京安酷智芯科技有限公司 | Noise reduction method, circuit system, electronic equipment and computer readable storage medium |
| CN113766150B (en) * | 2021-08-31 | 2024-03-26 | 北京安酷智芯科技有限公司 | Noise reduction method, circuit system, electronic equipment and computer readable storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109978774B (en) | 2021-06-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109978774A (en) | Multiframe continuously waits the denoising fusion method and device of exposure images | |
| CN111539879B (en) | Video blind denoising method and device based on deep learning | |
| US7570832B2 (en) | Image clean-up and pre-coding | |
| JP6094863B2 (en) | Image processing apparatus, image processing method, program, integrated circuit | |
| TWI596573B (en) | Image processing device for reducing image noise and the method thereof | |
| JP4118059B2 (en) | Method and apparatus for digital video processing | |
| CN105850114A (en) | Method for inverse tone mapping of an image | |
| CN109963048B (en) | Noise reduction method, noise reduction device and noise reduction circuit system | |
| CN109743473A (en) | Video image 3 D noise-reduction method, computer installation and computer readable storage medium | |
| CN109767408B (en) | Image processing method, image processing device, storage medium and computer equipment | |
| JP5822157B2 (en) | Noise reduction apparatus, noise reduction method, and program | |
| CN107993189B (en) | Image tone dynamic adjustment method and device based on local blocking | |
| CN111709904B (en) | Image fusion method and device | |
| DE102020200310A1 (en) | Method and system for reducing haze for image processing | |
| RU2419880C2 (en) | Method and apparatus for calculating and filtering disparity map based on stereo images | |
| CN105631854A (en) | FPGA platform-based self-adaptive image definition evaluation algorithm | |
| CN104751417A (en) | Color noise reducing method, device and image processing system | |
| CN113379609B (en) | Image processing method, storage medium and terminal equipment | |
| CN105069764B (en) | A kind of image de-noising method and system based on Edge track | |
| CN109859138B (en) | Infrared image enhancement method based on human visual characteristics | |
| JP5559012B2 (en) | Image processing apparatus and control method thereof | |
| US11669939B1 (en) | Burst deblurring with kernel estimation networks | |
| CN114240767A (en) | Image wide dynamic range processing method and device based on exposure fusion | |
| CN111260590B (en) | Image noise reduction method and related product | |
| Munezawa et al. | Noise removal method for moving images using 3-D and time-domain total variation regularization decomposition |
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 |