CN110111268A - The removing rain based on single image method and device learnt based on dark and blurred width - Google Patents
The removing rain based on single image method and device learnt based on dark and blurred width Download PDFInfo
- Publication number
- CN110111268A CN110111268A CN201910314127.3A CN201910314127A CN110111268A CN 110111268 A CN110111268 A CN 110111268A CN 201910314127 A CN201910314127 A CN 201910314127A CN 110111268 A CN110111268 A CN 110111268A
- Authority
- CN
- China
- Prior art keywords
- rain
- image
- defogging
- channel
- blurred width
- 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 21
- 230000000694 effects Effects 0.000 claims abstract description 36
- 238000012549 training Methods 0.000 claims abstract description 21
- 238000012545 processing Methods 0.000 claims abstract description 13
- 238000006243 chemical reaction Methods 0.000 claims abstract description 9
- 238000000926 separation method Methods 0.000 claims abstract description 7
- 238000012360 testing method Methods 0.000 claims abstract description 5
- 238000003672 processing method Methods 0.000 claims description 7
- 239000003086 colorant Substances 0.000 claims description 5
- 239000000654 additive Substances 0.000 claims description 4
- 230000000996 additive effect Effects 0.000 claims description 4
- 235000013399 edible fruits Nutrition 0.000 claims 2
- 230000002411 adverse Effects 0.000 description 6
- 239000002245 particle Substances 0.000 description 5
- 239000003595 mist Substances 0.000 description 4
- 230000002194 synthesizing effect Effects 0.000 description 3
- 239000000443 aerosol Substances 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 230000008030 elimination Effects 0.000 description 2
- 238000003379 elimination reaction Methods 0.000 description 2
- 230000002708 enhancing effect Effects 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 239000000203 mixture Substances 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000010521 absorption reaction Methods 0.000 description 1
- 230000000052 comparative effect Effects 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 230000007850 degeneration Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 239000000428 dust Substances 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000002401 inhibitory effect Effects 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000011084 recovery Methods 0.000 description 1
- 239000004576 sand Substances 0.000 description 1
- 238000009738 saturating Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 230000000007 visual effect Effects 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
- G06T5/77—Retouching; Inpainting; Scratch removal
-
- 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
- G06T2207/20081—Training; Learning
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
Abstract
The present invention relates to a kind of removing rain based on single image method and devices learnt based on dark and blurred width, and wherein method includes: step S1: being pre-processed to original image defogging;Step S2: image carries out low-and high-frequency separation after defogging is pre-processed, and takes high frequency section to carry out the conversion of color space, is converted to YCbCr color space from RGB color;Step S3: the input that the channel Y of the corresponding YCbCr color space of the image of training set part learns as blurred width is subjected to model training;Step S4: the input that the channel Y of test set part is learnt as blurred width, obtain the channel Y without rain figure;Step S5: it goes rain figure and low pass basal layer to combine on high pass layer and obtains preliminary removing rain effect picture;Step S6: based on image after defogging pretreatment, going rain effect picture to optimize processing preliminary, obtains final removing rain effect picture.Compared with prior art, the present invention has many advantages, such as that color rendition degree is high.
Description
Technical field
The present invention relates to image processing techniques, more particularly, to a kind of single width figure learnt based on dark and blurred width
As rain removing method and device.
Background technique
Computer vision system is widely used in various industries, including the tracking of video monitoring, vision and navigation, intelligence are handed over
Logical, entertainment industry etc..Computer vision system under indoor situations has been commonly used and has studied, and some outdoor items
Part, such as rain, snow, mist are still the challenging problem of computer vision system.Common adverse weather is according to composition
Particle and visual signature are broadly divided into stable state adverse weather (referring mainly to mist, haze) and dynamic adverse weather (refers mainly to rain, snow, sand
Dust storm etc.).Wherein, the aerosol systems structure that stable state adverse weather is mainly made of particles such as very small water droplet and dusts
At.Absorption and scattering process due to aerosol particle to atmosphere light cause image pixel intensities variation in image relatively slow.With it is steady
State adverse weather is compared, and the composition particle of the dynamics adverse weather such as rainwater weather is usually much greater, and these particles are fast
Its motion profile is easy to be influenced by extraneous factors such as wind while speed falls, and hides to generate fuzzy and rain line in the picture
Gear causes the foundation of model and scene recovery to become more complicated, causes the problems such as regional area is excessively bright, background image is fuzzy.
The degeneration of rainy day picture quality greatly constrains the function of the outdoor intelligent vision system such as vision monitoring, vision guided navigation and target following
Can, and raindrop graininess is changeable, the rainy line direction of different situations and thickness are all different.Therefore, how research is from all kinds of
The image that high quality is recovered in rainy day degraded image has high research and application value.
Summary of the invention
It is an object of the present invention to overcome the above-mentioned drawbacks of the prior art and provide one kind based on dark and
The removing rain based on single image method and device of blurred width study.
The purpose of the present invention can be achieved through the following technical solutions:
A kind of removing rain based on single image processing method learnt based on dark and blurred width, comprising:
Step S1: original image defogging is pre-processed;
Step S2: image carries out low-and high-frequency separation after defogging is pre-processed, and high frequency section is taken to carry out the conversion of color space,
YCbCr color space is converted to from RGB color;
Step S3: the channel Y of the corresponding YCbCr color space of the image of training set part is learnt as blurred width
Input carries out model training;
Step S4: the channel Y of the corresponding YCbCr color space of the image of test set part is learnt as blurred width
Input, obtain the channel Y without rain figure;
Step S5: synthesizing high pass layer without rain figure and Cb channel C r channel for the channel Y and go rain figure, and will be in step S2
High pass layer is gone rain figure and low pass basal layer to combine and obtains preliminary going rain effect as low pass basal layer by the low frequency part arrived
Figure;
Step S6: it based on image after defogging pretreatment, goes rain effect picture to optimize processing preliminary, obtains final
Remove rain effect picture.
The pretreated result of defogging in the step S1 are as follows:
Wherein: Fr (x) is the image after defogging pretreatment, and F (x) is original image, and A is that global atmosphere is bright, tr (x)
For transmissivity, T0 is a threshold value for being set as 0.1.
The step S6 is specifically included:
Step S61: image detail part is extracted in image after defogging pretreatment using Gauss high-pass filter;
Step S62: the detail section transparency extracted is adjusted to setting ratio;
Step S63: the detail section after adjustment transparency is superimposed to as secondary colour and is gone as the preliminary of primary colours
It obtains final removing rain effect picture on rain effect picture.
The setting ratio is 30%, the detail section after adjusting transparency specifically:
IdeT=(Ide×Alpha+127)/255
Wherein: IdeTFor the detail section after adjustment transparency, Alpha is transparency, IdeFor original detail section.
The mathematic(al) representation of additive process in the step S63 are as follows:
Wherein: B removes rain effect picture, I for finalpreTo go what rain levels of detail combined with basal layer tentatively to remove rain
Effect picture.
A kind of removing rain based on single image processing unit learnt based on dark and blurred width, including memory, processor,
And the program for being stored in memory and being executed by the processor, the processor realize following step when executing described program
It is rapid:
Step S1: original image defogging is pre-processed;
Step S2: image carries out low-and high-frequency separation after defogging is pre-processed, and high frequency section is taken to carry out the conversion of color space,
YCbCr color space is converted to from RGB color;
Step S3: the channel Y of the corresponding YCbCr color space of the image of training set part is learnt as blurred width
Input carries out model training;
Step S4: the channel Y of the corresponding YCbCr color space of the image of test set part is learnt as blurred width
Input, obtain the channel Y without rain figure;
Step S5: synthesizing high pass layer without rain figure and Cb channel C r channel for the channel Y and go rain figure, and will be in step S2
High pass layer is gone rain figure and low pass basal layer to combine and obtains preliminary going rain effect as low pass basal layer by the low frequency part arrived
Figure;
Step S6: it based on image after defogging pretreatment, goes rain effect picture to optimize processing preliminary, obtains final
Remove rain effect picture.
Compared with prior art, the invention has the following advantages:
1), because the fog that rainwater generates is handled completely, will can thus make hidden in picture using the principle of dark defogging
It ensconces fog moderate rain line to highlight, to promote subsequent rainwater removal effect.
2) image after defogging is passed through high-pass filtering and low-pass filtering operation, respectively obtains levels of detail and base
Plinth layer, only the channel Y by high frequency section after color space conversion carries out blurred width study and carries out rain handling, and one will
The color image training problem of triple channel is converted to single channel levels of detail training problem, considerably reduces net training time,
To improve network efficiency;Two remain more raw informations, so that color rendition degree is strengthened.
3) with wave filter to original mist elimination image extract image detail, finally by detail view with 30 percent it is saturating
Lightness as secondary colour be superimposed to as primary colours tentatively go on rain effect picture go rain figure to get to eventually pass through enhancing.
Detailed description of the invention
Fig. 1 is the key step flow diagram of the method for the present invention;
Fig. 2 is the comparative examples figure that the method for the present invention generates effect picture and other algorithms generate effect picture.
Specific embodiment
The present invention is described in detail with specific embodiment below in conjunction with the accompanying drawings.The present embodiment is with technical solution of the present invention
Premised on implemented, the detailed implementation method and specific operation process are given, but protection scope of the present invention is not limited to
Following embodiments.
A kind of removing rain based on single image processing method learnt based on dark and blurred width, this method pass through computer journey
The form of sequence realizes that the computer system includes memory, processor, and is stored in memory simultaneously by computer system
The program executed by processor, as shown in Figure 1, processor performs the steps of when executing program
Step S1: pre-processing original image defogging,
Specifically, removing the misty rain scape of distant place in image using the algorithm of the preferential defogging of dark.For any
Input fog free images FD, dark channel prior is just like giving a definition:
Wherein: FrDIndicate each channel of color image RGB, Ω (x) represents the window centered on pixel x.Then
We can be in conjunction with image after the graph model solution defogging that mists:
F (x)=Fr (x) tr (x)+A (1-tr (x))
Wherein: F (x) is input to mist elimination image, and Fr (x) is fog free images to be restored, and A is global atmosphere light,
Tr (x) is transmissivity.Finally after a series of deformation of formula, our fogless solving model are as follows:
Wherein: Fr (x) is the image after defogging pretreatment, and F (x) is original image, and A is that global atmosphere is bright, tr (x)
For transmissivity, T0 is a threshold value for being set as 0.1.
Step S2: image carries out low-and high-frequency separation after defogging is pre-processed, and high frequency section is taken to carry out the conversion of color space,
YCbCr color space is converted to from RGB color, specifically, image after defogging is carried out low-and high-frequency separation by the application, takes height
Frequency part carries out the conversion of color space, is converted to YCbCr color space from RGB color.Finally high frequency section is taken to exist
The input that the channel Y in YCbCr color space learns as blurred width.The specific mathematic(al) representation of RGB color conversion
For;
Y=0.257*R+0.564*G+0.098*B+16
Cb=-0.148*R-0.291*G+0.439*B+128
Cr=0.439*R-0.368*G-0.071*B+128
Wherein: wherein R, G, B are respectively the numerical value in three kinds of channels of RGB color.
Step S3: the channel Y of the corresponding YCbCr color space of the image of training set part is learnt as blurred width
Input carries out model training, i.e. input data is: X=(x1,x2,...,xN)T∈RN×M, the application is using single order TS model, originally
Application will be inputted using single order TS fuzzy model:
xs=(xs1, xs2..., sxm)
It maps that in i-th of sub-fuzzy system, and this sub-fuzzy system has kiA fuzzy rule.It is remembered
Are as follows:
Wherein:It is the coefficient generated at random between [0,1], k=1,2 ..., kiIt is the fuzzy of i-th of fuzzy system
Rule.
The intermediate output valve Zsi of input for enhancement layer, s-th of training sample of i-th of sub-fuzzy system can be indicated
Are as follows:
Wherein:For fuzzy rule coefficient
The intermediate Zp exported that p sub-fuzzy system is mapped to enhancement layer can be indicated are as follows:
Wherein: j is the number of sub-fuzzy system.Then to ZjNonlinear transformation is carried out to get the output of enhancement layer, table is arrived
It is shown as:
Finally, the output O of whole network, i.e. the channel Y is without rain figure are as follows:
O=Fp+HgWe
Wherein: FpIt indicates that p-th of sub-fuzzy system, W are coefficient matrix, is calculated by pseudoinverse
Step S5: synthesizing high pass layer without rain figure and Cb channel C r channel for the channel Y and go rain figure, and will be in step S2
High pass layer is gone rain figure and low pass basal layer to combine and obtains preliminary going rain effect as low pass basal layer by the low frequency part arrived
Figure;
Step S6: it based on image after defogging pretreatment, goes rain effect picture to optimize processing preliminary, obtains final
Rain effect picture is removed, is specifically included:
Step S61: image detail part is extracted in image after defogging pretreatment using Gauss high-pass filter;
Step S62: the detail section transparency extracted being adjusted to setting ratio, it is preferred that the setting ratio is 30%,
Detail section after adjusting transparency specifically:
IdeT=(Ide×Alpha+127)/255
Wherein: IdeTFor the detail section after adjustment transparency, Alpha is transparency, IdeFor original detail section.
Step S63: the detail section after adjustment transparency is superimposed to as secondary colour and is gone as the preliminary of primary colours
It obtains final removing rain effect picture on rain effect picture.
The mathematic(al) representation of additive process in step S63 are as follows:
Wherein: B removes rain effect picture, I for finalpreTo go what rain levels of detail combined with basal layer tentatively to remove rain
Effect picture.
Since many deep learning rain removing methods are directly trained without pretreatment, lead to network training accuracy decline.
And since that there are parameters is various for existing deep learning method, the problems such as training iteration time is too long, and amount of training data is big, therefore
The input that blurred width learns is converted to the height in the channel Y in YCbCr space after highlighting nearby apparent rain line by the application
Frequently, network training convergence rate is greatly improved.Finally for details smooth, herein using the side of details mixing superposition
Method enhances web results, as shown in Fig. 2, enhancing image detail in the case where inhibiting rain line.Therefore it is higher having
Accuracy rate while, training speed is significantly promoted, it is specific as shown in table 1:
Claims (10)
1. a kind of removing rain based on single image processing method learnt based on dark and blurred width characterized by comprising
Step S1: original image defogging is pre-processed;
Step S2: image carries out low-and high-frequency separation after defogging is pre-processed, and high frequency section is taken to carry out the conversion of color space, from
RGB color is converted to YCbCr color space;
Step S3: the input that the channel Y of the corresponding YCbCr color space of the image of training set part is learnt as blurred width
Carry out model training;
Step S4: learn using the channel Y of the corresponding YCbCr color space of the image of test set part as blurred width defeated
Enter, obtain the channel Y without rain figure;
Step S5: the channel Y is synthesized into high pass layer without rain figure and the channel Cb channel C r and goes rain figure, and will be obtained in step S2
High pass layer is gone rain figure and low pass basal layer to combine and obtains preliminary removing rain effect picture as low pass basal layer by low frequency part;
Step S6: based on image after defogging pretreatment, going rain effect picture to optimize processing preliminary, obtains final removing rain
Effect picture.
2. a kind of removing rain based on single image processing method learnt based on dark and blurred width according to claim 1,
It is characterized in that, the pretreated result of defogging in the step S1 are as follows:
Wherein: Fr (x) is the image after defogging pretreatment, and F (x) is original image, and A is that global atmosphere is bright, and tr (x) is
Rate is penetrated, T0 is a threshold value for being set as 0.1.
3. a kind of removing rain based on single image processing method learnt based on dark and blurred width according to claim 1,
It is characterized in that, the step S6 is specifically included:
Step S61: image detail part is extracted in image after defogging pretreatment using Gauss high-pass filter;
Step S62: the detail section transparency extracted is adjusted to setting ratio;
Step S63: using adjust transparency after detail section as secondary colour be superimposed to as primary colours tentatively go rain to imitate
It obtains final removing rain effect picture on fruit figure.
4. a kind of removing rain based on single image processing method learnt based on dark and blurred width according to claim 3,
It is characterized in that, the setting ratio is 30%, the detail section after adjusting transparency specifically:
IdeT=(Ide×Alpha+127)/255
Wherein: IdeTFor the detail section after adjustment transparency, Alpha is transparency, IdeFor original detail section.
5. a kind of removing rain based on single image processing method learnt based on dark and blurred width according to claim 4,
It is characterized in that, the mathematic(al) representation of the additive process in the step S63 are as follows:
Wherein: B removes rain effect picture, I for finalpreTo go what rain levels of detail combined with basal layer tentatively to go rain effect
Figure.
6. a kind of removing rain based on single image processing unit learnt based on dark and blurred width, which is characterized in that including storage
Device, processor, and the program for being stored in memory and being executed by the processor, when the processor executes described program
It performs the steps of
Step S1: original image defogging is pre-processed;
Step S2: image carries out low-and high-frequency separation after defogging is pre-processed, and high frequency section is taken to carry out the conversion of color space, from
RGB color is converted to YCbCr color space;
Step S3: the input that the channel Y of the corresponding YCbCr color space of the image of training set part is learnt as blurred width
Carry out model training;
Step S4: learn using the channel Y of the corresponding YCbCr color space of the image of test set part as blurred width defeated
Enter, obtain the channel Y without rain figure;
Step S5: the channel Y is synthesized into high pass layer without rain figure and the channel Cb channel C r and goes rain figure, and will be obtained in step S2
High pass layer is gone rain figure and low pass basal layer to combine and obtains preliminary removing rain effect picture as low pass basal layer by low frequency part;
Step S6: based on image after defogging pretreatment, going rain effect picture to optimize processing preliminary, obtains final removing rain
Effect picture.
7. a kind of removing rain based on single image processing unit learnt based on dark and blurred width according to claim 6,
It is characterized in that, the pretreated result of defogging in the step S1 are as follows:
Wherein: Fr (x) is the image after defogging pretreatment, and F (x) is original image, and A is that global atmosphere is bright, and tr (x) is
Rate is penetrated, T0 is a threshold value for being set as 0.1.
8. a kind of removing rain based on single image processing unit learnt based on dark and blurred width according to claim 6,
It is characterized in that, the step S6 is specifically included:
Step S61: image detail part is extracted in image after defogging pretreatment using Gauss high-pass filter;
Step S62: the detail section transparency extracted is adjusted to setting ratio;
Step S63: using adjust transparency after detail section as secondary colour be superimposed to as primary colours tentatively go rain to imitate
It obtains final removing rain effect picture on fruit figure.
9. a kind of removing rain based on single image processing unit learnt based on dark and blurred width according to claim 8,
It is characterized in that, the setting ratio is 30%, the detail section after adjusting transparency specifically:
IdeT=(Ide×Alpha+127)/255
Wherein: IdeTFor the detail section after adjustment transparency, Alpha is transparency, IdeFor original detail section.
10. a kind of removing rain based on single image processing unit learnt based on dark and blurred width according to claim 9,
It is characterized in that, the mathematic(al) representation of the additive process in the step S63 are as follows:
Wherein: B removes rain effect picture, I for finalpreTo go what rain levels of detail combined with basal layer tentatively to go rain effect
Figure.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910314127.3A CN110111268B (en) | 2019-04-18 | 2019-04-18 | Single image rain removing method and device based on dark channel and fuzzy width learning |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910314127.3A CN110111268B (en) | 2019-04-18 | 2019-04-18 | Single image rain removing method and device based on dark channel and fuzzy width learning |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110111268A true CN110111268A (en) | 2019-08-09 |
CN110111268B CN110111268B (en) | 2021-08-03 |
Family
ID=67485815
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910314127.3A Active CN110111268B (en) | 2019-04-18 | 2019-04-18 | Single image rain removing method and device based on dark channel and fuzzy width learning |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110111268B (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110544217A (en) * | 2019-08-30 | 2019-12-06 | 深圳市商汤科技有限公司 | image processing method and device, electronic equipment and storage medium |
CN110928187A (en) * | 2019-12-03 | 2020-03-27 | 北京工业大学 | Sewage treatment process fault monitoring method based on fuzzy width self-adaptive learning model |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20130236116A1 (en) * | 2012-03-08 | 2013-09-12 | Industrial Technology Research Institute | Method and apparatus for single-image-based rain streak removal |
CN105046653A (en) * | 2015-06-12 | 2015-11-11 | 中国科学院深圳先进技术研究院 | Method and system for removing raindrops in videos |
CN105184761A (en) * | 2015-08-28 | 2015-12-23 | 中国科学院深圳先进技术研究院 | Image rain removing method based on wavelet analysis and system |
CN105931208A (en) * | 2015-08-22 | 2016-09-07 | 首都师范大学 | Physical model-based low-illuminance image enhancement algorithm |
CN107909556A (en) * | 2017-11-27 | 2018-04-13 | 天津大学 | Video image rain removing method based on convolutional neural networks |
CN109064419A (en) * | 2018-07-12 | 2018-12-21 | 四川大学 | A kind of removing rain based on single image method based on WLS filtering and multiple dimensioned sparse expression |
-
2019
- 2019-04-18 CN CN201910314127.3A patent/CN110111268B/en active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20130236116A1 (en) * | 2012-03-08 | 2013-09-12 | Industrial Technology Research Institute | Method and apparatus for single-image-based rain streak removal |
CN105046653A (en) * | 2015-06-12 | 2015-11-11 | 中国科学院深圳先进技术研究院 | Method and system for removing raindrops in videos |
CN105931208A (en) * | 2015-08-22 | 2016-09-07 | 首都师范大学 | Physical model-based low-illuminance image enhancement algorithm |
CN105184761A (en) * | 2015-08-28 | 2015-12-23 | 中国科学院深圳先进技术研究院 | Image rain removing method based on wavelet analysis and system |
CN107909556A (en) * | 2017-11-27 | 2018-04-13 | 天津大学 | Video image rain removing method based on convolutional neural networks |
CN109064419A (en) * | 2018-07-12 | 2018-12-21 | 四川大学 | A kind of removing rain based on single image method based on WLS filtering and multiple dimensioned sparse expression |
Non-Patent Citations (3)
Title |
---|
DANA BERMAN 等: "Non-Local Image Dehazing", 《2016 IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION》 * |
SHUANG FENG 等: "Fuzzy Broad Learning System: A Novel Neuro-Fuzzy Model for Regression and Classification", 《IEEE TRANSACTIONS ON CYBERNETICS》 * |
李耀维: "单幅雨天图像去雨方法研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110544217A (en) * | 2019-08-30 | 2019-12-06 | 深圳市商汤科技有限公司 | image processing method and device, electronic equipment and storage medium |
CN110928187A (en) * | 2019-12-03 | 2020-03-27 | 北京工业大学 | Sewage treatment process fault monitoring method based on fuzzy width self-adaptive learning model |
CN110928187B (en) * | 2019-12-03 | 2021-02-26 | 北京工业大学 | Sewage treatment process fault monitoring method based on fuzzy width self-adaptive learning model |
Also Published As
Publication number | Publication date |
---|---|
CN110111268B (en) | 2021-08-03 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109272455B (en) | Image defogging method based on weak supervision generation countermeasure network | |
CN105354806B (en) | Rapid defogging method and system based on dark | |
CN105046666B (en) | A kind of method of the real-time defogging of traffic video based on dark primary priori | |
CN107451966B (en) | Real-time video defogging method implemented by guiding filtering through gray level image | |
US20160334546A1 (en) | Weather recognition method and device based on image information detection | |
CN110148093B (en) | Image defogging improvement method based on dark channel prior | |
CN106251301A (en) | A kind of single image defogging method based on dark primary priori | |
CN105225210A (en) | A kind of self-adapting histogram based on dark strengthens defogging method capable | |
US20170132764A1 (en) | Image inpainting system and method for using the same | |
CN107103591A (en) | A kind of single image to the fog method based on image haze concentration sealing | |
CN106600560A (en) | Image defogging method for automobile data recorder | |
CN104867121B (en) | Image Quick demisting method based on dark primary priori and Retinex theories | |
CN110310241A (en) | A kind of more air light value traffic image defogging methods of fusion depth areas segmentation | |
CN110490155B (en) | Method for detecting unmanned aerial vehicle in no-fly airspace | |
CN104809709A (en) | Single-image self-adaptation defogging method based on domain transformation and weighted quadtree decomposition | |
CN109523474A (en) | A kind of enhancement method of low-illumination image based on greasy weather degradation model | |
CN106296618A (en) | A kind of color image defogging method based on Gaussian function weighted histogram regulation | |
CN106204494A (en) | A kind of image defogging method comprising large area sky areas and system | |
CN110211052A (en) | A kind of single image to the fog method based on feature learning | |
CN110111268A (en) | The removing rain based on single image method and device learnt based on dark and blurred width | |
CN105809643A (en) | Image enhancing method based on self-adaptive block channel stretching | |
CN106251296A (en) | A kind of image defogging method and system | |
CN109118450A (en) | A kind of low-quality images Enhancement Method under the conditions of dust and sand weather | |
CN111598814B (en) | Single image defogging method based on extreme scattering channel | |
CN115063434B (en) | Low-low-light image instance segmentation method and system based on feature 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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |