CN109410146A - A kind of image deblurring algorithm based on Bi-Skip-Net - Google Patents
A kind of image deblurring algorithm based on Bi-Skip-Net Download PDFInfo
- Publication number
- CN109410146A CN109410146A CN201811298475.8A CN201811298475A CN109410146A CN 109410146 A CN109410146 A CN 109410146A CN 201811298475 A CN201811298475 A CN 201811298475A CN 109410146 A CN109410146 A CN 109410146A
- Authority
- CN
- China
- Prior art keywords
- feature
- skip
- layer
- shallow
- sampling
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 claims abstract description 15
- 238000005070 sampling Methods 0.000 claims description 24
- 238000013461 design Methods 0.000 claims description 8
- 230000006835 compression Effects 0.000 claims description 7
- 238000007906 compression Methods 0.000 claims description 7
- 230000009467 reduction Effects 0.000 claims description 6
- 208000037170 Delayed Emergence from Anesthesia Diseases 0.000 claims description 5
- 238000011176 pooling Methods 0.000 claims description 3
- 230000000694 effects Effects 0.000 abstract description 16
- 238000013135 deep learning Methods 0.000 abstract description 9
- 238000012545 processing Methods 0.000 abstract description 4
- 238000011084 recovery Methods 0.000 description 8
- 230000006870 function Effects 0.000 description 5
- 230000007246 mechanism Effects 0.000 description 5
- 230000004069 differentiation Effects 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 230000000052 comparative effect Effects 0.000 description 3
- 238000007796 conventional method Methods 0.000 description 3
- 238000013527 convolutional neural network Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000003475 lamination Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 238000006467 substitution reaction Methods 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
- G06T5/73—Deblurring; Sharpening
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The present invention relates to digital image processing fields, especially a kind of image deblurring method based on Bi-Skip-Net, it is a kind of Bi-Skip-Net network to realize blur image restoration, it is intended to the problems such as solving high time complexity existing for existing deep learning deblurring algorithm, reconstruction inaccuracy, restored image there are grid effects.A kind of Bi-Skip-Net network disclosed by the invention as GAN (Generative Adversarial Network) generation network, aim to solve the problem that existing deep learning deblurring algorithm there are the shortcomings that, by comparing existing optimal algorithm, the present invention improves 0.1s on time complexity, can go up in image complex pattern originality and averagely improve 1dB.
Description
Technical field
The present invention relates to digital image processing field, especially a kind of image deblurring method based on Bi-Skip-Net,
This method is to realize blur image restoration by Bi-Skip-Net network.
Technical background
Deblurring technology is image and the theme that field of video processing is widely studied.Based on fuzzy caused by camera shake
The image quality of image, vision perception are seriously affected in a sense.As one, image preprocessing field and its important
Branch, the promotion of deblurring technology directly affect the performance of other computer vision algorithms makes, such as foreground segmentation, object detection, row
For analysis etc.；It also affects the coding efficiency of image simultaneously.Therefore, a kind of high performance deblurring algorithm gesture is studied must
Row.
Document 1-3 describes the deblurring technology of image and video processing, deep learning deblurring algorithm；Document 1:
Kupyn O, Budzan V,Mykhailych M,et al.DeblurGAN:Blind Motion Deblurring Using
Conditional Adversarial Networks[J].arXiv preprint arXiv:1711.07064,2017.Document
2:Nah S, Kim T H, Lee K M.Deep multi-scale convolutional neural network for
dynamic scene deblurring[C]//CVPR.2017,1(2):3.Document 3:Sun J, Cao W, Xu Z, et
al.Learning a convolutional neural network for non-uniform motion blur
removal[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern
Recognition.2015:769-777.。
In general, image deblurring algorithm can be divided into traditional algorithm based on probabilistic model and based on deep learning
Deblurring algorithm.Traditional algorithm explains that the fuzzy origin cause of formation, the process of camera shake can be mapped as fuzzy core using convolution model
Track PSF (Point Spread Function).Clear image is restored in the case where fuzzy core is unknown, this problem belongs to
Fixed (ill-posed) problem of discomfort recycles the fuzzy core of assessment to carry out so needing first ambiguous estimation core on ordinary meaning
It returns convolution operation and obtains restored image.Deblurring algorithm based on deep learning obtains the latent of image using deep layer network structure
In information, and then realize blur image restoration.Fuzzy kernel estimates and non-blind deconvolution may be implemented in deep learning deblurring algorithm
Two operate to carry out image restoration, while can also be using generation confrontation mechanism come restored image.This patent aims to solve the problem that
Fuzzy algorithmic approach there are the shortcomings that:
1) time complexity is high,
2) reconstruction inaccuracy,
3) there are grid effects for restored image.
Summary of the invention
The invention proposes a kind of Bi-Skip-Net networks as GAN (Generative Adversarial
Network generation network), it is intended to solve the disadvantage that existing deep learning deblurring algorithm exists.It is existing optimal by comparing
Algorithm, the present invention improve 0.1s on time complexity, can go up in image complex pattern originality and averagely improve 1dB.
Technical solution provided by the invention is as follows: (note: needing to be illustrated technical solution with natural language, Bu Nengyong
Narrating mode " as figure ", method and technology scheme preferably by: write in the way of step 1: step 2 ...)
The present invention realizes blur image restoration using confrontation network mechanism is generated, and devises a kind of Bi-Skip-Net
Network is as generator therein.Specific step is as follows:
1): input blurred picture, by convolution kernel having a size of 7x7, the convolutional layer that step-length is 1 obtains shallow-layer feature；
2): shallow-layer feature is obtained into the depth characteristic under current scale by 3 residual blocks；
3): adding the mode of residual error to obtain the shallow-layer feature under next scale depth characteristic progress down-sampling；
4): according to the down-sampling frequency n of regulation, repeating step 2,3 obtain shallow-layer feature and depth spy under different scale
Sign, and depth characteristic is not obtained under smallest dimension；
5): using the shallow-layer feature of smallest dimension as essential characteristic；
6): it is 1x1 that the shallow-layer feature of upper one layer of scale, which is passed through convolution kernel size, and the convolutional layer that step-length is 1 obtains shallow-layer
Dimensionality reduction feature；It is 3x3 that corresponding depth characteristic, which is passed through convolution kernel size, and the convolutional layer that step-length is 2 obtains depth dimensionality reduction feature,
And it connects and is up-sampled with essential characteristic；Feature after up-sampling is connected to obtain current scale with shallow-layer dimensionality reduction feature
Under essential characteristic；
7): repeating step 6 until sampling operation ends；
8): it is 7x7 that obtained essential characteristic, which is passed through convolution kernel size, and the convolutional layer that step-length is 1 obtains residual error feature；
9): being added residual error feature to obtain restored image with input picture；
…
Using the mode of Bi-Skip-Net plus residual error as generator.
In step 4), the down-sampling number according to regulation is 5.
Wherein, blurred picture obtains restored image by generator, and the task of differentiation is to distinguish restored image as far as possible
And clear image；And the task of generator is the ability for cheating arbiter as far as possible to reduce the differentiation to two kinds of images.
The Bi-Skip-Net network consists of three parts: the path contract (D), the path Skip (S) and
The path expand (U).Contract layers of progress down-sampling realize Feature Compression, and Skip layers for connecting further feature and shallow-layer
Feature, expand layers are up-sampled.Wherein D*, S*, U* are the feature under corresponding down-sampling scale.
Characteristic manipulation under the sampling scale, in the path contract, current signature passes through 3 residual blocks
(3xResBlock) obtains further feature, and the residual error mode being added using pond (pooling) with convolution is next to obtain
The feature of scale；In the path Skip, by the convolution of 1x1 come to compression shallow-layer feature, by the convolution of 3x3 come compression depth
Feature；It in the path expand, realizes that feature is connected by concat, and feature up-sampling is realized by the deconvolution of 3x3.
The present invention has the following technical effect that since present invention employs Bi-Skip-Net networks as GAN
The generation network of (Generative Adversarial Network) has following technical effect that compared with prior art
1, time complexity is low；Compare conventional method, it is traditional to go motion blur method using fuzzy kernel estimates and non-blind
Two steps of deconvolution, and the two steps be both needed to carry out successive ignition can be only achieved preferable recovery effect, just because of this,
Cause the time for handling individual motion blur image longer；And the model that the present invention designs can cause to avoid successive ignition optimization
Time loss.
2, reconstruction is accurate；Conventional method is compared, in conventional methods where, fuzzy kernel estimates inaccuracy, which will cause, restored
The Fault recovery of image information in journey, and the operation of non-blind warp often will cause texture part and ringing effect occurs；The present invention is set
The twin spans connection network of meter is all extracted depth characteristic and shallow-layer feature on each layer of scale, is connected by feature, network exists
It can restore more detailed information to a certain degree.
3, grid effect is not present in restored image, compares existing deep learning method, existing most deep learning sides
Method realized in upper sampling process using warp lamination, and since each deconvolution all has certain sawtooth effect, this makes
Obtaining last restored image, there is also some sawtooth, i.e., the grid effects of the invention mentioned.
Design and principle for a better understanding of the present invention carry out the present invention detailed with reference to the accompanying drawings and examples
Thin description.But the description of specific embodiment is not in any way limit the scope of the present invention.
Detailed description of the invention
Fig. 1 is that generation of the invention fights network mechanism；
Fig. 2 is Bi-Skip-Net network structure of the invention；
Fig. 3 is the characteristic manipulation under a sampling scale of the invention；
Fig. 4 Generator Design: Bi-Skip-Net+ residual error；
Fig. 5 a-d is the subjective comparative of the present invention with other algorithms；Wherein,
Fig. 5 a: blurred picture；
The recovery effect of Fig. 5 b:Nah et al.；
The recovery effect of Fig. 5 c:Kupyn et al.；
The recovery effect of Fig. 5 d:Bi-Skip-Net.
Specific embodiment
Fig. 1 is that the generation that the present invention uses fights network mechanism.Wherein, blurred picture obtains restored map by generator
Picture, the task of differentiation are to distinguish restored image and clear image as far as possible；And the task of generator is to cheat arbiter as far as possible
To reduce the ability of the differentiation to two kinds of images.
Specific step is as follows for the embodiment of the present invention:
(1) generator and arbiter are designed, principle is as shown in figure 4, be that the blurred picture of building passes through Bi-Skip-Net
Generator, and obtain clearly building picture；Other any blurred pictures can generate clearly picture with this model.
(2) network is trained using following loss function,
WhereinTo fight loss function,For conditional loss function, λ is the weight of conditional loss function.
Pass through maximizationTo optimize arbiter D；
Optimize generator G by minimum formula 3；
WhereinIt designs as follows:
Wherein, L, S respectively indicate model in the output and true value of different levels, and α value is 1 or 2, entire conditional loss letter
Number is standardized by port number c, width w and height h.
(3) using trained network as final restoration model.
As shown in Figure 1, the method for the embodiment of the present invention realizes blur image restoration using confrontation network mechanism is generated.Figure
2 be Bi-Skip-Net network structure, takes network structure shown in Fig. 2, devises a kind of Bi-Skip-Net network to make
For generator therein.
Arbiter parameter in the Bi-Skip-Net network structure is shown in Table 1. arbiter parameter lists.
1. arbiter parameter list of table
| # | Layer | Parameter dimensions | Step-length |
| 1 | conv | 32x3x5x5 | 2 |
| 2 | conv | 64x32x5x5 | 1 |
| 3 | conv | 64x64x5x5 | 2 |
| 4 | conv | 128x64x5x5 | 1 |
| 5 | conv | 128x128x5x5 | 4 |
| 6 | conv | 256x128x5x5 | 1 |
| 7 | conv | 256x256x5x5 | 4 |
| 8 | conv | 512x256x5x5 | 1 |
| 9 | conv | 512x512x4x4 | 4 |
| 10 | fc | 512x1x1x1 | - |
As shown in Fig. 2, the Bi-Skip-Net network of design of the embodiment of the present invention consists of three parts: the path contract
Including D0, D1, D2 and D3 (D),；The path Skip (S), including S0, S1, S2 and S3；And the path expand (U), including U0,
U1, U2 and U3.Contract layers of progress down-sampling realize Feature Compression, and Skip layers are used to connect further feature and shallow-layer feature,
Expand layers are up-sampled.Wherein D* (D0, D1, D2 and D3), S* (S0, S1, S2 and S3), U* (U0, U1, U2 and U3) are
Feature under corresponding down-sampling scale.
Fig. 3 is the characteristic manipulation under sampling scale, as shown in figure 3, in the path contract, i.e. compressed path, when
Preceding feature obtains further feature by 3 residual blocks (3xResBlock), and be added using pond (pooling) with convolution
Residual error mode obtains the feature of next scale；In the path Skip, i.e. mid-span path, by the convolution of 1x1 come shallow to compressing
Layer feature, compresses further feature by the convolution of 3x3；Come in the path expand by concat, i.e. concatenate
It realizes feature connection, and feature up-sampling is realized by the deconvolution of 3x3.
Fig. 4 Generator Design: Bi-Skip-Net+ residual error, as shown in figure 4, finally adding residual error using Bi-Skip-Net
Mode is as generator.
The present invention is implemented with the comparing results of other algorithms see Table 2 for details the present invention and other algorithms on GoPro data set
Test comparison.
2. present invention of table and test comparison of other algorithms on GoPro data set
Fig. 5 a-d is the subjective comparative of the present invention with other algorithms.Fig. 5 a is blurred picture, and Fig. 5 b is the recovery of Nah et al.
Effect, Fig. 5 c are the recovery effect of Kupyn et al., and Fig. 5 d is the recovery effect of Bi-Skip-Net of the invention.The picture lower left corner
Text " HARDWARE " be beyond recognition or recognize fuzzy in other three pictures, the present invention clearly can be restored and be identified.From
It is obvious to the repairing effect of blurred picture that people subjective comparative can be seen that the present invention.
It should be noted that the purpose for publicizing and implementing example is to help to further understand the present invention, but the skill of this field
Art personnel, which are understood that, not to be departed from the present invention and spirit and scope of the appended claims, and various substitutions and modifications are all
It is possible.Therefore, the present invention should not be limited to embodiment disclosure of that, and the scope of protection of present invention is with claim
Subject to the range that book defines.
Claims (5)
1. a kind of image deblurring method based on Bi-Skip-Net, includes the following steps:
1) blurred picture is inputted, by convolution kernel having a size of 7x7, the convolutional layer that step-length is 1 obtains shallow-layer feature；
2) shallow-layer feature is obtained into the depth characteristic under current scale by 3 residual blocks；
3) mode of residual error is added to obtain the shallow-layer feature under next scale depth characteristic progress down-sampling；
4) according to the down-sampling frequency n of regulation, step 2 is repeated, 3 obtain the shallow-layer feature and depth characteristic under different scale, and
And depth characteristic is not obtained under smallest dimension；
5) using the shallow-layer feature of smallest dimension as essential characteristic；
It 6) is 1x1 by convolution kernel size by the shallow-layer feature of upper one layer of scale, it is special that the convolutional layer that step-length is 1 obtains shallow-layer dimensionality reduction
Sign；It is 3x3 that corresponding depth characteristic, which is passed through convolution kernel size, and the convolutional layer that step-length is 2 obtains depth dimensionality reduction feature, and and base
Eigen series connection is up-sampled；Feature after up-sampling is connected to obtain the base under current scale with shallow-layer dimensionality reduction feature
Eigen；
7) step 6 is repeated until sampling operation ends；
It 8) is 7x7 by convolution kernel size by obtained essential characteristic, the convolutional layer that step-length is 1 obtains residual error feature；
9) it is added residual error feature to obtain restored image with input picture；
10) add the mode of residual error as generator using Bi-Skip-Net.
2. image deblurring method according to claim 1, it is characterised in that:
Step 4) is 5 according to the down-sampling number of regulation.
3. image deblurring method according to claim 1, it is characterised in that:
The Bi-Skip-Net network consists of three parts: the path contract (D), the path Skip (S) and the road expand
Diameter (U)；Contract layers of progress down-sampling realize Feature Compression, and Skip layers are used to connect further feature and shallow-layer feature,
Expand layers are up-sampled；Wherein D*, S*, U* are the feature under corresponding down-sampling scale.
4. image deblurring method according to claim 3, it is characterised in that:
Characteristic manipulation under the sampling scale, in the path contract, current signature passes through 3 residual blocks
(3xResBlock) obtains further feature, and the residual error mode being added using pond (pooling) with convolution is next to obtain
The feature of scale；In the path Skip, by the convolution of 1x1 come to compression shallow-layer feature, by the convolution of 3x3 come compression depth spy
Sign；It in the path expand, realizes that feature is connected by concat, and feature up-sampling is realized by the deconvolution of 3x3.
5. image deblurring method according to claim 1, it is characterised in that:
Generator described in step 10) designs as follows,
1. network is trained using following loss function,
WhereinTo fight loss function,For conditional loss function, λ is the weight of conditional loss function；
Pass through maximizationTo optimize arbiter D；
Optimize generator G by minimum formula 3；
WhereinIt designs as follows:
Wherein, L, S respectively indicate model in the output and true value of different levels, and α value is 1 or 2, entire conditional loss function quilt
Port number c, width w and height h are standardized；
2. using trained network as final restoration model.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811298475.8A CN109410146A (en) | 2018-11-02 | 2018-11-02 | A kind of image deblurring algorithm based on Bi-Skip-Net |
| PCT/CN2018/117634 WO2020087607A1 (en) | 2018-11-02 | 2018-11-27 | Bi-skip-net-based image deblurring method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811298475.8A CN109410146A (en) | 2018-11-02 | 2018-11-02 | A kind of image deblurring algorithm based on Bi-Skip-Net |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109410146A true CN109410146A (en) | 2019-03-01 |
Family
ID=65471437
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811298475.8A Pending CN109410146A (en) | 2018-11-02 | 2018-11-02 | A kind of image deblurring algorithm based on Bi-Skip-Net |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN109410146A (en) |
| WO (1) | WO2020087607A1 (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110570375A (en) * | 2019-09-06 | 2019-12-13 | 腾讯科技（深圳）有限公司 | image processing method, image processing device, electronic device and storage medium |
| CN111612711A (en) * | 2019-05-31 | 2020-09-01 | 北京理工大学 | Improved picture deblurring method based on generation countermeasure network |
| CN112102184A (en) * | 2020-09-04 | 2020-12-18 | 西北工业大学 | Image deblurring method based on Scale-Encoder-Decoder-Net network |
| CN113570516A (en) * | 2021-07-09 | 2021-10-29 | 湖南大学 | Image blind motion deblurring method based on CNN-Transformer hybrid self-encoder |
Families Citing this family (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111986102B (en) * | 2020-07-15 | 2024-02-27 | 万达信息股份有限公司 | Digital pathological image deblurring method |
| CN112070658B (en) * | 2020-08-25 | 2024-04-16 | 西安理工大学 | Deep learning-based Chinese character font style migration method |
| CN112070693B (en) * | 2020-08-27 | 2024-03-26 | 西安理工大学 | Single dust image recovery method based on gray world adaptive network |
| CN112184590B (en) * | 2020-09-30 | 2024-03-26 | 西安理工大学 | Single dust image recovery method based on gray world self-guiding network |
| CN112330554B (en) * | 2020-10-30 | 2024-01-19 | 西安工业大学 | Structure learning method for deconvolution of astronomical image |
| CN112561819A (en) * | 2020-12-17 | 2021-03-26 | 温州大学 | Self-filtering image defogging algorithm based on self-supporting model |
| CN113538263A (en) * | 2021-06-28 | 2021-10-22 | 江苏威尔曼科技有限公司 | Motion blur removing method, medium, and device based on improved DeblurgAN model |
| CN113592736B (en) * | 2021-07-27 | 2024-01-12 | 温州大学 | Semi-supervised image deblurring method based on fused attention mechanism |
| CN113947589A (en) * | 2021-10-26 | 2022-01-18 | 北京理工大学 | Missile-borne image deblurring method based on countermeasure generation network |
| CN114119395B (en) * | 2021-11-15 | 2024-06-11 | 北京理工大学 | Image processing system and method integrating distortion detection and restoration |
| CN114240771A (en) * | 2021-11-23 | 2022-03-25 | 无锡学院 | Image deblurring system and method based on dual control network |
| CN114723630A (en) * | 2022-03-31 | 2022-07-08 | 福州大学 | Image deblurring method and system based on cavity double-residual multi-scale depth network |
| CN114913095B (en) * | 2022-06-08 | 2024-03-12 | 西北工业大学 | Depth deblurring method based on domain adaptation |
| CN114841897B (en) * | 2022-06-08 | 2024-03-15 | 西北工业大学 | Depth deblurring method based on self-adaptive fuzzy kernel estimation |
| CN115760589A (en) * | 2022-09-30 | 2023-03-07 | 浙江大学 | Image optimization method and device for motion blurred image |
| CN117058038B (en) * | 2023-08-28 | 2024-04-30 | 北京航空航天大学 | Diffraction blurred image restoration method based on even convolution deep learning |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106251303A (en) * | 2016-07-28 | 2016-12-21 | 同济大学 | A kind of image denoising method using the degree of depth full convolutional encoding decoding network |
| US20170365046A1 (en) * | 2014-08-15 | 2017-12-21 | Nikon Corporation | Algorithm and device for image processing |
| CN107689034A (en) * | 2017-08-16 | 2018-02-13 | 清华-伯克利深圳学院筹备办公室 | A kind of training method of neutral net, denoising method and device |
| CN108629743A (en) * | 2018-04-04 | 2018-10-09 | 腾讯科技（深圳）有限公司 | Processing method, device, storage medium and the electronic device of image |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8737687B2 (en) * | 2011-12-30 | 2014-05-27 | Honeywell International Inc. | System and method for tracking a subject using raw images and tracking errors |
| CN108460742A (en) * | 2018-03-14 | 2018-08-28 | 日照职业技术学院 | A kind of image recovery method based on BP neural network |
| CN108711141B (en) * | 2018-05-17 | 2022-02-15 | 重庆大学 | Motion blurred image blind restoration method using improved generation type countermeasure network |
-
2018
- 2018-11-02 CN CN201811298475.8A patent/CN109410146A/en active Pending
- 2018-11-27 WO PCT/CN2018/117634 patent/WO2020087607A1/en active Application Filing
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20170365046A1 (en) * | 2014-08-15 | 2017-12-21 | Nikon Corporation | Algorithm and device for image processing |
| CN106251303A (en) * | 2016-07-28 | 2016-12-21 | 同济大学 | A kind of image denoising method using the degree of depth full convolutional encoding decoding network |
| CN107689034A (en) * | 2017-08-16 | 2018-02-13 | 清华-伯克利深圳学院筹备办公室 | A kind of training method of neutral net, denoising method and device |
| CN108629743A (en) * | 2018-04-04 | 2018-10-09 | 腾讯科技（深圳）有限公司 | Processing method, device, storage medium and the electronic device of image |
Non-Patent Citations (3)
| Title |
|---|
| KAI LIN 等: "Adaptive Integration Skip Compensation Neural Networks for Removing Mixed Noise in Image", 《PCM2018:ADVANCES IN MULTIMEDIA INFORMATION PROCESSING》 * |
| OREST KUPYN 等: "DeblurGAN: Blind Motion Deblurring Using Conditional Adversarial Networks", 《ARXIV》 * |
| WEIWEI FU 等: "Arrears Prediction For Electricity Customer Through Wgan-Gp", 《2017 IEEE 2ND INFORMATION TECHNOLOGY, NETWORKING, ELECTRONIC AND AUTOMATION CONTROL CONFERENCE（ITNEC）》 * |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111612711A (en) * | 2019-05-31 | 2020-09-01 | 北京理工大学 | Improved picture deblurring method based on generation countermeasure network |
| CN111612711B (en) * | 2019-05-31 | 2023-06-09 | 北京理工大学 | Picture deblurring method based on generation of countermeasure network improvement |
| CN110570375A (en) * | 2019-09-06 | 2019-12-13 | 腾讯科技（深圳）有限公司 | image processing method, image processing device, electronic device and storage medium |
| CN110570375B (en) * | 2019-09-06 | 2022-12-09 | 腾讯科技（深圳）有限公司 | Image processing method, device, electronic device and storage medium |
| CN112102184A (en) * | 2020-09-04 | 2020-12-18 | 西北工业大学 | Image deblurring method based on Scale-Encoder-Decoder-Net network |
| CN113570516A (en) * | 2021-07-09 | 2021-10-29 | 湖南大学 | Image blind motion deblurring method based on CNN-Transformer hybrid self-encoder |
| CN113570516B (en) * | 2021-07-09 | 2022-07-22 | 湖南大学 | Image blind motion deblurring method based on CNN-Transformer hybrid self-encoder |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2020087607A1 (en) | 2020-05-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109410146A (en) | A kind of image deblurring algorithm based on Bi-Skip-Net | |
| Dong et al. | Model-guided deep hyperspectral image super-resolution | |
| CN109785258B (en) | Face image restoration method based on multi-discriminator generated countermeasure network | |
| Thai et al. | Image classification using support vector machine and artificial neural network | |
| Cui et al. | Deep network cascade for image super-resolution | |
| CN109544475A (en) | Bi-Level optimization method for image deblurring | |
| Li et al. | Face hallucination based on sparse local-pixel structure | |
| Huang et al. | Deep hyperspectral image fusion network with iterative spatio-spectral regularization | |
| CN109543548A (en) | A kind of face identification method, device and storage medium | |
| CN105590304B (en) | Super-resolution image reconstruction method and device | |
| Dong et al. | Robust tensor approximation with Laplacian scale mixture modeling for multiframe image and video denoising | |
| CN111696038A (en) | Image super-resolution method, device, equipment and computer-readable storage medium | |
| Li et al. | Lightweight single image super-resolution with dense connection distillation network | |
| Hua et al. | Dynamic scene deblurring with continuous cross-layer attention transmission | |
| CN113344110B (en) | Fuzzy image classification method based on super-resolution reconstruction | |
| CN103020940A (en) | Local feature transformation based face super-resolution reconstruction method | |
| Xu et al. | Multi-scale dehazing network via high-frequency feature fusion | |
| CN110503608A (en) | The image de-noising method of convolutional neural networks based on multi-angle of view | |
| CN104252715B (en) | Single line image-based three-dimensional reconstruction method | |
| CN117315752A (en) | Training method, device, equipment and medium for face emotion recognition network model | |
| Kligvasser et al. | Deep self-dissimilarities as powerful visual fingerprints | |
| Jeevan et al. | WaveMixSR: Resource-efficient neural network for image super-resolution | |
| Gao et al. | Motion deblurring algorithm for wind power inspection images based on Ghostnet and SE attention mechanism | |
| CN115861810A (en) | Remote sensing image change detection method and system based on multi-head attention and self-supervision learning | |
| Jia et al. | Learning rich information for quad bayer remosaicing and denoising |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190301 |