CN102768760A - Quick image dehazing method on basis of image textures - Google Patents
Quick image dehazing method on basis of image textures Download PDFInfo
- Publication number
- CN102768760A CN102768760A CN2012102299292A CN201210229929A CN102768760A CN 102768760 A CN102768760 A CN 102768760A CN 2012102299292 A CN2012102299292 A CN 2012102299292A CN 201210229929 A CN201210229929 A CN 201210229929A CN 102768760 A CN102768760 A CN 102768760A
- Authority
- CN
- China
- Prior art keywords
- image
- pixel
- input picture
- scope
- matrix
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Abstract
The invention provides a quick image dehazing method on the basis of image textures, which includes the following steps: step I, estimating the atmospheric environment light of an input image I (x); step II, estimating the transmission matrix t (x) of the input image I (x); and step III, regaining the dehazed image as per the atmospheric environment light and the transmission matrix t (x). In the step II, according to the edge detection result of the input image, a size-changeable sliding block is used for obtaining the rough transmission matrix, namely the smaller sliding block is used for precise processing at regions where the depth of field changes suddenly, and a larger sliding block is used at regions where the depth of field changes slightly, so that the operation speed is improved. The result obtained through the changeable sliding block is more precise than that obtained through a fixed block, and the result of the rough transmission matrix can be directly used for post processing.
Description
Technical field
The invention belongs to image processing techniques, particularly the image defogging method capable.
Background technology
Under the greasy weather condition, the image that Outdoor Scene is taken tends to receive the influence of impurity particle in the atmosphere, and it is unclear to cause color distortion, contrast reduction, object to thicken, and has influenced the result of use of vision systems such as outdoor scene video monitoring and Target Recognition.Airborne suspended particle, for example: dust, mist etc., can make body surface true reflected light generation chromatic dispersion and decay, caused the distortion of object color.And these suspended particles can reflect atmospheric environment light, and the atmospheric environment light component reduces picture contrast after getting into vision system.Mist elimination is widely used in Flame Image Process and computer vision field, and important Research Significance and practical value are arranged.At first, after image or video do not have the mist processing, can significantly improve the visuality of object in the scene, can also correct because the color displacement that the atmospheric environment light component brings makes color truer.Secondly, the algorithm that most computer vision is handled often supposes that input picture is the reflected light of body surface, and therefore, the performance of these algorithms inevitably can receive color displacement, the influence that contrast reduces.At last, mist elimination can produce the depth information of image, can be applied to fields such as scene understanding.
Disposal route for Misty Image mainly is divided into two types at present: Misty Image Enhancement Method and Misty Image restored method.The method scope of application that Misty Image strengthens is wider, often from the angle of Flame Image Process, improves the contrast of image, can improve the visual effect of image, the details of outstanding image, but the picture contrast after recovering sometimes is higher, not necessarily has authenticity.It is the physical process that the research Misty Image degrades that Misty Image is restored, and sets up atmosphere degeneration physical model, utilizes view data estimation model parameter, estimates the recovery image according to physical model afterwards.This method is with strong points, and the mist elimination effect that obtains has certain authenticity, can obtain the depth information of image, does not generally have information loss, and the key point of processing is the estimation of parameter in the model.
To the single width Misty Image, existing Misty Image restoration methods is often based on the prior imformation rule of Misty Image.Typical method has Tan method, Fattal method and He method.The Tan method finds that no mist image is often high than the mist picture contrast is arranged, and based on this prior imformation rule, makes the contrast of the regional area of mist image reach maximum; Image after can obtaining to recover; Visual effect after the recovery is better, but contrast is often very high, and not necessarily has authenticity.The prior imformation of Fattal method is that the reflected light component of hypothesis surround lighting component and body surface is incoherent, and this method image restored has certain authenticity, but can not good treatment thick fog image.The He method is based on the prior imformation of helping secretly, and the picture quality that recovers is truer, but this computing is complicated, calculated amount is big, can not accomplish image is handled in real time, has restricted its application in reality.
Summary of the invention
Technical matters to be solved by this invention is to provide a kind of rapid speed, image to recover the higher Misty Image restoration methods of quality.
The present invention is adopted technical scheme to be by solving the problems of the technologies described above, and the quick defogging method capable of a kind of image based on image texture may further comprise the steps:
The atmospheric environment light of step 1, estimation input picture I (x);
The transmission matrix t (x) of step 2, estimation input picture I (x):
1) respectively three passages of R, G, B of input picture I (x) are asked for corresponding matrix Ira, Iga, Iba:
Ira=I
R(x)/A
R
Iga=I
G(x)/A
G
Iba=I
B(x)/A
B
Wherein, I
R(x), I
G(x), I
B(x) be respectively R, G, the B component image of input picture I (x), A
R, A
G, A
BBe respectively R, G, the B component of said atmospheric environment light;
2) input picture I (x) is carried out rim detection, obtain the input picture edge detection graph, and with marginal point and non-marginal point binaryzation, marginal point is put 1, non-marginal point puts 0;
3) to matrix Ira, Iga, Iba carry out sliding shoe to be handled, and obtains rough transmission matrix t' (x):
Wherein, Ω (x) is for being the sliding shoe at center with pixel x, and the size of sliding shoe is decided through the concrete condition of edge pixel o'clock in the first scope fritter and the second scope fritter, and first scope is greater than second scope;
In the first scope fritter that with pixel x is the center, do not have the edge pixel point, then the size of sliding shoe Ω (x) is first scope;
In the first scope fritter that with pixel x is the center, have the edge pixel point, and be not have the edge pixel point in the second scope fritter at center with pixel x, then the size of sliding shoe Ω (x) is second scope;
In the first scope fritter that with pixel x is the center, there is the edge pixel point; And be to have the edge pixel point in the second scope fritter at center with pixel x; Judge then whether pixel x is fringe node; In this way, then the size of sliding shoe Ω (x) is 1 * 1, otherwise the size of sliding shoe Ω (x) is the size that does not comprise the largest block of any marginal point in second scope;
4) rough transmission matrix t' (x) is carried out medium filtering and obtain meticulous transmission matrix t (x);
Step 3, recover the image behind the mist elimination according to atmospheric environment light and transmission matrix t (x).
The present invention can obtain transmission matrix fast through the method for step 2.Edge detection results according to input picture; Use the sliding shoe piece of variable size to obtain rough transmission matrix: to adopt less sliding shoe accurately to handle in the place of depth of field sudden change (edge pixel point is many); The place that changes little (edge pixel point is few) in the depth of field adopts sliding shoe greatly to handle, and improves arithmetic speed.The variable-block process result also can directly be carried out subsequent treatment with this rough transmission matrix result than meticulous with fixing piece process result.
Further, the concrete implementation method of step 1 is following:
1) handle through input picture I (x) being carried out sliding shoe, what obtain input picture I (x) helps Dark (x) secretly:
Wherein, Ω (x) is for being the sliding shoe at center with pixel x, and the sliding shoe size is 15 * 15;
2) in helping Dark (x) secretly, seek brightness greater than all pixels of preset luminance threshold and preserve its position in image in matrix L o;
3) convert input picture I (x) to gray level image Gray (x), the location of pixels correspondence of finding out matrix L o storage in gray level image Gray (x) pixel value and preserve this pixel value in Matrix C;
4) find out the highest pixel value of brightness and preserve the position LL of respective pixel in image in Matrix C;
5) pixel value of the position LL of input picture I (x) is an atmospheric environment light.
The chosen area of surround lighting during the preset luminance threshold of the present invention's employing in estimating atmospheric environment light is selected to help secretly, the speed that makes atmospheric environment light obtain is fast, helps real-time processing.
Concrete, the concrete grammar that recovers the image J (x) behind the mist elimination according to atmospheric environment light and transmission matrix t (x) is:
Wherein, A is an atmospheric environment light, t
0Be preset noise threshold, I (x) is an input picture, and J (x) is the image behind the mist elimination, and t (x) is a transmission matrix.
Further, better in order to recover image effect, increased brightness adjustment coefficient t in the algorithm of the image J (x) after recovering mist elimination
1Improve the brightness of image J (x).
The invention has the beneficial effects as follows that the picture quality that recovers is higher, calculated amount is little, and rapid speed can reach the purpose of real-time processing, can be applied in the real system easily.
Description of drawings
Fig. 1: schematic flow sheet of the present invention.
Fig. 2: input mist image I (x) arranged.
Fig. 3: input picture help figure secretly.
Fig. 4: the edge detection graph of input picture.
Fig. 5: rough transmission matrix figure t ' (x).
Fig. 6: the transmission matrix figure t (x) after the optimization.
Fig. 7: the design sketch J (x) behind the mist elimination.
Embodiment
Present embodiment uses the simulated program of matlab2010b exploitation.At double-core 2.5GCPU, on the PC platform of 2G internal memory, the image of 240*320 pixel as shown in Figure 2 to be carried out mist elimination handle, concrete steps are as shown in Figure 1.
Step 1, estimation atmospheric environment light A.Specifically comprise step by step following:
The 1st step: handle through input picture being carried out sliding shoe, what obtain input picture I (x) helps Dark (x) secretly, can obtain the figure that helps secretly of input picture as shown in Figure 3.Input picture I (x) helps Dark (x) secretly and can obtain by following formula:
At this moment, sliding shoe Ω (x) is to be 15 * 15 the fritter at center with pixel x.
The 2nd step: in helping Dark (x) secretly, seek the pixel of brightness, and preserve its position in image in matrix L o greater than preset luminance threshold.
The 3rd step: convert I (x) to gray level image Gray (x), can obtain the input picture gray-scale map.Taking-up is in the pixel at Lo place, position in Gray (x), and is stored in the Matrix C.Convert I (x) to gray level image, conversion method is following: I (x) RGB model conversion is become the HSI model, and the luminance component I among the HSI is exactly the gray level image that I (x) converts to.
The 4th step: in Matrix C, find out the highest pixel of brightness, preserve its position in image in LL.
The 5th step: be in the pixel of position LL at input picture I (x), be the atmospheric environment light A of acquisition.Pixel LL is atmospheric environment light among the I (x).
Step 2, estimation Misty Image transmission matrix.Specifically comprise step by step following:
The 1st step: respectively three passages of input picture I (x) RGB are asked for matrix:
Ira=I
R(x)/A
R
Iga=I
G(x)/A
G
Iba=I
B(x)/A
B
Three components of RGB that three formula are respectively I (x) are divided by three components of RGB of atmospheric environment light A.
The 2nd step: input picture I (x) is carried out rim detection, obtain input picture edge detection graph as shown in Figure 4, and with marginal point and non-marginal point binaryzation.Input picture I (x) is carried out rim detection, and what rim detection was used is the sobel operator.Binaryzation refers to, and marginal point is put 1, and non-marginal point puts 0.
The 3rd step: obtain rough transmission matrix t' (x), rough transmission matrix figure promptly as shown in Figure 5.With the local fritter Ω (x) at pixel x place, to matrix Ira, Iga, Iba carry out sliding shoe to be handled, and obtains rough transmission matrix t' (x).T' (x) can obtain according to the following equation:
Do with pixel x be the center 15 * 15 local fritter Ω ' (x), do with pixel x be the center make 5 * 5 local fritter Ω into " (x).For pixel x, the size of local fritter Ω (x) selects to divide following situation discussion:
If when there was not any marginal point in (x) in Ω ', Ω (x) was taken as Ω ' (x).
If when there was marginal point in Ω ' in (x), Ω (x) value was following:
" when not having any marginal point (x), Ω (x) value Ω " (x) as Ω.
" when existing marginal point and center pixel x to be marginal point (x), Ω (x) is taken as pixel x as Ω.
" have marginal point (x), but marginal point is when central element, Ω (x) is taken as at Ω " is the largest block that does not comprise any marginal point at center with pixel x (x) as Ω.
The 4th step: obtain to optimize the meticulous transmission matrix t (x) in back, the transmission matrix figure after promptly optimizing, as shown in Figure 6.Transmission coefficient matrix t' (x) is carried out medium filtering, can obtain meticulous transmission matrix t (x).
Step 3, acquisition recover back image J (x), and be as shown in Figure 7.Recovering image J (x) obtains according to following formula:
T wherein
0Value 0.3, t
1Value 1.0-1.3.
Experimental result representes that embodiment approximately needs about 3s handling every width of cloth image, and processing speed is fast, can requirement of real time.
Claims (6)
1. the quick defogging method capable of the image based on image texture is characterized in that, may further comprise the steps:
The atmospheric environment light of step 1, estimation input picture I (x);
The transmission matrix t (x) of step 2, estimation input picture I (x):
1) respectively three passages of R, G, B of input picture I (x) are asked for corresponding matrix Ira, Iga, Iba:
Ira=I
R(x)/A
R
Iga=I
G(x)/A
G
Iba=I
B(x)/A
B
Wherein, I
R(x), I
G(x), I
B(x) be respectively R, G, the B component image of input picture I (x), A
R, A
G, A
BBe respectively R, G, the B component of said atmospheric environment light;
2) input picture I (x) is carried out rim detection, obtain the input picture edge detection graph, and with marginal point and non-marginal point binaryzation, marginal point is put 1, non-marginal point puts 0;
3) to matrix Ira, Iga, Iba carry out sliding shoe to be handled, and obtains rough transmission matrix t' (x):
Wherein, Ω (x) is for being the sliding shoe at center with pixel x, and the size of sliding shoe is decided through the concrete condition of edge pixel o'clock in the first scope fritter and the second scope fritter, and first scope is greater than second scope;
In the first scope fritter that with pixel x is the center, do not have the edge pixel point, then the size of sliding shoe Ω (x) is first scope;
In the first scope fritter that with pixel x is the center, have the edge pixel point, and be not have the edge pixel point in the second scope fritter at center with pixel x, then the size of sliding shoe Ω (x) is second scope;
In the first scope fritter that with pixel x is the center, there is the edge pixel point; And be to have the edge pixel point in the second scope fritter at center with pixel x; Judge then whether pixel x is fringe node; In this way, then the size of sliding shoe Ω (x) is 1 * 1, otherwise the size of sliding shoe Ω (x) is the size that does not comprise the largest block of any marginal point in second scope;
4) rough transmission matrix t' (x) is carried out medium filtering and obtain meticulous transmission matrix t (x);
Step 3, recover the image behind the mist elimination according to atmospheric environment light and transmission matrix t (x).
2. the quick defogging method capable of a kind of according to claim 1 image based on image texture is characterized in that, the concrete grammar that recovers the image J (x) behind the mist elimination according to atmospheric environment light and transmission matrix t (x) is:
Wherein, A is an atmospheric environment light, t
0Be preset noise threshold, I (x) is an input picture, and J (x) is the image behind the mist elimination, and t (x) is a transmission matrix.
3. the quick defogging method capable of a kind of according to claim 1 image based on image texture is characterized in that, the concrete grammar that recovers the image J (x) behind the mist elimination according to atmospheric environment light and transmission matrix t (x) is:
Wherein, A is an atmospheric environment light, t
0Be preset noise threshold, I (x) is an input picture, and J (x) is the image behind the mist elimination, and t (x) is a transmission matrix, t
1Be brightness adjustment coefficient.
4. like the quick defogging method capable of the said a kind of image of claim 3, it is characterized in that t based on image texture
0Value 0.3, t
1Span is 1.0 to 1.3.
5. the quick defogging method capable of a kind of according to claim 1 image based on image texture is characterized in that, the said first scope block sizes is that 15 * 15, the second scope block sizes is 5 * 5.
6. the quick defogging method capable of a kind of according to claim 1 image based on image texture is characterized in that, the concrete steps of the atmospheric environment light of estimation input picture I (x) are following:
1) handle through input picture I (x) being carried out sliding shoe, what obtain input picture I (x) helps Dark (x) secretly:
Wherein, Ω (x) is for being the sliding shoe at center with pixel x, and the sliding shoe size is 15 * 15;
2) in helping Dark (x) secretly, seek brightness greater than all pixels of preset luminance threshold and preserve its position in image in matrix L o;
3) convert input picture I (x) to gray level image Gray (x), the location of pixels correspondence of finding out matrix L o storage in gray level image Gray (x) pixel value and preserve this pixel value in Matrix C;
4) find out the highest pixel value of brightness and preserve the position LL of respective pixel in image in Matrix C;
5) pixel value of the position LL of input picture I (x) is an atmospheric environment light.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201210229929.2A CN102768760B (en) | 2012-07-04 | 2012-07-04 | Quick image dehazing method on basis of image textures |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201210229929.2A CN102768760B (en) | 2012-07-04 | 2012-07-04 | Quick image dehazing method on basis of image textures |
Publications (2)
Publication Number | Publication Date |
---|---|
CN102768760A true CN102768760A (en) | 2012-11-07 |
CN102768760B CN102768760B (en) | 2014-12-17 |
Family
ID=47096153
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201210229929.2A Expired - Fee Related CN102768760B (en) | 2012-07-04 | 2012-07-04 | Quick image dehazing method on basis of image textures |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN102768760B (en) |
Cited By (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102982513A (en) * | 2012-12-04 | 2013-03-20 | 电子科技大学 | Adaptive image defogging method based on textures |
CN103136735A (en) * | 2013-03-07 | 2013-06-05 | 中国科学院光电技术研究所 | Single image defogging method based on dual-scale dark channel |
CN103150708A (en) * | 2013-01-18 | 2013-06-12 | 上海交通大学 | Image quick defogging optimized method based on black channel |
CN103500440A (en) * | 2013-09-27 | 2014-01-08 | 中国人民解放军空军工程大学 | Method for eliminating cloud and haze of atmospheric degraded image |
CN103500460A (en) * | 2013-09-18 | 2014-01-08 | 大连理工大学 | Image fog effect simulation method based on depth |
CN103700079A (en) * | 2013-12-17 | 2014-04-02 | 中国科学院深圳先进技术研究院 | Image defogging method and device |
CN103714519A (en) * | 2013-12-03 | 2014-04-09 | 中国民用航空总局第二研究所 | Rapid parallel image defogging method |
CN103903234A (en) * | 2014-03-12 | 2014-07-02 | 南京第五十五所技术开发有限公司 | Real-time image defogging method based on image field depth |
CN103985086A (en) * | 2013-02-07 | 2014-08-13 | 株式会社日立制作所 | Image demisting device, camera device and image demisting method |
CN104200445A (en) * | 2014-09-26 | 2014-12-10 | 常熟理工学院 | Image defogging method with optimal contrast ratio and minimal information loss |
CN104240193A (en) * | 2013-09-15 | 2014-12-24 | 西南科技大学 | Singe image defogging method based on luminance component minimization |
CN104281999A (en) * | 2013-07-12 | 2015-01-14 | 东北师范大学 | Single image defogging method based on structural information |
CN104318519A (en) * | 2014-09-26 | 2015-01-28 | 南京邮电大学 | Edge substitution method-based image haze removal method |
CN105741248A (en) * | 2016-02-02 | 2016-07-06 | 深圳市未来媒体技术研究院 | Method for removing hazed and degraded part in image |
CN110473155A (en) * | 2019-08-02 | 2019-11-19 | 西安工程大学 | A kind of image defogging method based on retina color-aware dark principle |
CN110738624A (en) * | 2019-10-18 | 2020-01-31 | 电子科技大学 | area self-adaptive image defogging system and method |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20110188775A1 (en) * | 2010-02-01 | 2011-08-04 | Microsoft Corporation | Single Image Haze Removal Using Dark Channel Priors |
US20110211758A1 (en) * | 2010-03-01 | 2011-09-01 | Microsoft Corporation | Multi-image sharpening and denoising using lucky imaging |
CN102411774A (en) * | 2011-08-08 | 2012-04-11 | 安防科技(中国)有限公司 | Processing method, device and system based on single-image defogging |
-
2012
- 2012-07-04 CN CN201210229929.2A patent/CN102768760B/en not_active Expired - Fee Related
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20110188775A1 (en) * | 2010-02-01 | 2011-08-04 | Microsoft Corporation | Single Image Haze Removal Using Dark Channel Priors |
US20110211758A1 (en) * | 2010-03-01 | 2011-09-01 | Microsoft Corporation | Multi-image sharpening and denoising using lucky imaging |
CN102411774A (en) * | 2011-08-08 | 2012-04-11 | 安防科技(中国)有限公司 | Processing method, device and system based on single-image defogging |
Non-Patent Citations (2)
Title |
---|
嵇晓强等: "暗原色先验图像去雾算法研究", 《光电子.激光》, vol. 22, no. 6, 30 June 2011 (2011-06-30), pages 926 - 930 * |
胡伟等: "基于暗通道优先的单幅图像去雾新方法", 《计算机研究与发展》, vol. 47, no. 12, 15 December 2010 (2010-12-15), pages 2132 - 2140 * |
Cited By (24)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102982513B (en) * | 2012-12-04 | 2015-08-19 | 电子科技大学 | A kind of adapting to image defogging method capable based on texture |
CN102982513A (en) * | 2012-12-04 | 2013-03-20 | 电子科技大学 | Adaptive image defogging method based on textures |
CN103150708A (en) * | 2013-01-18 | 2013-06-12 | 上海交通大学 | Image quick defogging optimized method based on black channel |
CN103150708B (en) * | 2013-01-18 | 2015-11-18 | 上海交通大学 | Based on the image Quick demisting optimization method of black channel |
CN103985086A (en) * | 2013-02-07 | 2014-08-13 | 株式会社日立制作所 | Image demisting device, camera device and image demisting method |
CN103136735A (en) * | 2013-03-07 | 2013-06-05 | 中国科学院光电技术研究所 | Single image defogging method based on dual-scale dark channel |
CN104281999A (en) * | 2013-07-12 | 2015-01-14 | 东北师范大学 | Single image defogging method based on structural information |
CN104240193A (en) * | 2013-09-15 | 2014-12-24 | 西南科技大学 | Singe image defogging method based on luminance component minimization |
CN104240193B (en) * | 2013-09-15 | 2017-02-15 | 西南科技大学 | Singe image defogging method based on luminance component minimization |
CN103500460A (en) * | 2013-09-18 | 2014-01-08 | 大连理工大学 | Image fog effect simulation method based on depth |
CN103500440A (en) * | 2013-09-27 | 2014-01-08 | 中国人民解放军空军工程大学 | Method for eliminating cloud and haze of atmospheric degraded image |
CN103714519A (en) * | 2013-12-03 | 2014-04-09 | 中国民用航空总局第二研究所 | Rapid parallel image defogging method |
CN103700079A (en) * | 2013-12-17 | 2014-04-02 | 中国科学院深圳先进技术研究院 | Image defogging method and device |
CN103903234A (en) * | 2014-03-12 | 2014-07-02 | 南京第五十五所技术开发有限公司 | Real-time image defogging method based on image field depth |
CN103903234B (en) * | 2014-03-12 | 2016-08-17 | 南京第五十五所技术开发有限公司 | A kind of real time imaging defogging method based on image depth |
CN104318519A (en) * | 2014-09-26 | 2015-01-28 | 南京邮电大学 | Edge substitution method-based image haze removal method |
CN104200445A (en) * | 2014-09-26 | 2014-12-10 | 常熟理工学院 | Image defogging method with optimal contrast ratio and minimal information loss |
CN104200445B (en) * | 2014-09-26 | 2017-04-26 | 常熟理工学院 | Image defogging method with optimal contrast ratio and minimal information loss |
CN104318519B (en) * | 2014-09-26 | 2017-07-07 | 南京邮电大学 | A kind of image defogging method based on edge method of substitution |
CN105741248A (en) * | 2016-02-02 | 2016-07-06 | 深圳市未来媒体技术研究院 | Method for removing hazed and degraded part in image |
CN105741248B (en) * | 2016-02-02 | 2019-02-22 | 深圳市未来媒体技术研究院 | A method of eliminating image haze |
CN110473155A (en) * | 2019-08-02 | 2019-11-19 | 西安工程大学 | A kind of image defogging method based on retina color-aware dark principle |
CN110738624A (en) * | 2019-10-18 | 2020-01-31 | 电子科技大学 | area self-adaptive image defogging system and method |
CN110738624B (en) * | 2019-10-18 | 2022-02-01 | 电子科技大学 | Area-adaptive image defogging system and method |
Also Published As
Publication number | Publication date |
---|---|
CN102768760B (en) | 2014-12-17 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN102768760B (en) | Quick image dehazing method on basis of image textures | |
CN102750674B (en) | Video image defogging method based on self-adapting allowance | |
CN101950416B (en) | Bidirectional filtration-based real-time image de-hazing and enhancing method | |
CN103218778B (en) | The disposal route of a kind of image and video and device | |
CN103020920B (en) | Method for enhancing low-illumination images | |
CN102063706B (en) | Rapid defogging method | |
CN103198459B (en) | Haze image rapid haze removal method | |
CN105225210A (en) | A kind of self-adapting histogram based on dark strengthens defogging method capable | |
CN104050637A (en) | Quick image defogging method based on two times of guide filtration | |
CN103077504B (en) | A kind of image defogging method capable based on self-adaptation illumination calculation | |
CN102831591A (en) | Gaussian filter-based real-time defogging method for single image | |
CN103020914B (en) | Based on the rapid image defogging method capable of spatial continuity principle | |
CN102665034A (en) | Night effect removal method for camera-collected video | |
CN103985091A (en) | Single image defogging method based on luminance dark priori method and bilateral filtering | |
CN105913390B (en) | A kind of image defogging method and system | |
CN103049888A (en) | Image/video demisting method based on combination of dark primary color of atmospheric scattered light | |
CN104272347A (en) | Image processing apparatus for removing haze contained in still image and method thereof | |
CN103914820A (en) | Image haze removal method and system based on image layer enhancement | |
CN103426151A (en) | Method and device for defogging image | |
CN103578083A (en) | Single image defogging method based on joint mean shift | |
CN103020921A (en) | Single image defogging method based on local statistical information | |
CN104966274A (en) | Local fuzzy recovery method employing image detection and area extraction | |
CN106657948A (en) | low illumination level Bayer image enhancing method and enhancing device | |
CN105139344A (en) | Method and system for removing influence of raindrops to single image based on frequency domain and phase consistency | |
CN104392417A (en) | Image haze removal method based on pixel dark channel and anisotropic diffusion filtering |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20141217 Termination date: 20170704 |
|
CF01 | Termination of patent right due to non-payment of annual fee |