CN107146209A - A kind of single image to the fog method based on gradient field - Google Patents
A kind of single image to the fog method based on gradient field Download PDFInfo
- Publication number
- CN107146209A CN107146209A CN201710301848.1A CN201710301848A CN107146209A CN 107146209 A CN107146209 A CN 107146209A CN 201710301848 A CN201710301848 A CN 201710301848A CN 107146209 A CN107146209 A CN 107146209A
- Authority
- CN
- China
- Prior art keywords
- mrow
- image
- gradient
- msub
- color
- 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 title claims abstract description 32
- 230000002708 enhancing effect Effects 0.000 claims abstract description 21
- 230000004927 fusion Effects 0.000 claims abstract description 16
- 238000011084 recovery Methods 0.000 claims description 11
- 238000012545 processing Methods 0.000 claims description 8
- 230000000694 effects Effects 0.000 claims description 5
- 230000006835 compression Effects 0.000 claims description 4
- 238000007906 compression Methods 0.000 claims description 4
- 238000002834 transmittance Methods 0.000 claims description 4
- 238000010276 construction Methods 0.000 claims description 3
- 238000006386 neutralization reaction Methods 0.000 claims description 3
- 230000008030 elimination Effects 0.000 description 3
- 238000003379 elimination reaction Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 239000003595 mist Substances 0.000 description 3
- 230000007547 defect Effects 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 239000000443 aerosol Substances 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 238000000205 computational method Methods 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 230000007850 degeneration Effects 0.000 description 1
- 230000003412 degenerative effect Effects 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 230000000593 degrading effect Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000003707 image sharpening Methods 0.000 description 1
- 239000012535 impurity Substances 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000007500 overflow downdraw method Methods 0.000 description 1
- 238000011160 research 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
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10024—Color image
-
- 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
- G06T2207/20024—Filtering details
- G06T2207/20032—Median filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20212—Image combination
- G06T2207/20221—Image fusion; Image merging
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
Abstract
The invention provides a kind of single image to the fog method based on gradient field, belong to image defogging technical field, solve that existing gradient field defogging method execution efficiency is low, result figure is partially dark, cross-color the problems such as.Fog-degraded image is transformed into gradient field first, by carrying out the enhancing of different scale to gradient area image, several gradient fields enhancing image is obtained;Then, the gradient stretch function based on brightness is constructed according to weber laws, and applied on multiple dimensioned, operation is modified to gradient field enhancing image;Finally the image after reconstruction is carried out to merge the image obtained after final defogging using the Weighted Fusion algorithm for recovering the factor with color.The present invention can effectively be improved and also increased compared with existing gradient field defogging method in the definition and color fidelity of fog-degraded image, execution efficiency.
Description
Technical field
The present invention relates to Digital Image Processing and computer vision field, especially a kind of single image based on gradient field
Defogging method.
Background technology
With the reduction and the lifting of performance of computer costs, video camera and computer vision system have been widely applied
In association areas such as target recognition and tracking, monitoring, intelligent navigations.However, the key of these system designs is IMAQ
Equipment can capture high-quality clearly image.Under the conditions of the greasy weather, the image of outdoor collection can be by the grain that suspended in air
, there are the phenomenons such as smudgy, contrast decline, cross-color in the influence of the impurity such as son, aerosol, and this is at follow-up image
The application of reason and computer vision system brings serious influence.Therefore, the degrading cause of Misty Image is probed into, and using figure
As the quality of defogging technology improvement fog-degraded image, the application to computer vision system has great importance.
Current image defogging algorithm is broadly divided into two classes:Method based on image enhaucament and the method based on image restoration.
Method based on image enhaucament is the feature from image, by using image enchancing method to the contrast of degraded image,
Color and brightness are handled to reach the purpose of defogging.The defect of this kind of method does not account for Misty Image degeneration
Reason, ignores the depth of view information of scene during image enhaucament so that enhanced image is in larger local right of the depth of field
It is more undesirable than degree and color.And the method based on image restoration is the causes for Degradation from fog-degraded image, physics is set up
Model, because the unknown parameter in model is more, it is impossible to directly recover clear fogless image, it is therefore desirable to extra letter
Breath, assumed condition or some prioris estimate corresponding parameter, and then inverting degenerative process, obtains the image after defogging.
In method based on image restoration most it is representational be Kaiming He propose the defogging method based on dark primary priori,
This method has obtained preferable defog effect, causes extensive concern.But there is long operational time, day dead zone in this method
Domain chromatic distortion, the problems such as more serious region contrast that degrades recovers unobvious.In recent years, for dark primary transcendental method and biography
The deficiency of system image enchancing method, Li Wujing (under Li Wu strength low visibilities, image sharpening technical research [D] Sichuan Universitys,
2012) a kind of algorithm for image enhancement based on gradient field is proposed, although avoid some defects of conventional algorithm, but the party
Method does not consider the monochrome information of original image when gradient field strengthens, and causes enhanced result figure partially dark, in addition in fusion
Operated during operation using gradient field Multiscale Fusion, maximum Grad is selected as the gradient finally merged, to noise-sensitive
And the number of times of Gradient Reconstruction is added, execution efficiency is relatively low.
Based on this, the invention provides a kind of single image to the fog method based on gradient field, examined when gradient field strengthens
The monochrome information of original image is considered, has introduced relative gradient, and constructed the gradient field function based on brightness, and answered
Use on multiple dimensioned, operation is modified to gradient field enhancing image so that enhancing result more meets human eye vision effect.This
Outside, in Multiscale Fusion, the image after reconstruction is melted present invention employs the Weighted Fusion algorithm recovered with color
Closing operation, it is contemplated that the ratio between each passage of original image, can effectively improve the color fidelity of result images, solution
Determined because each passage individually processing caused by cross-color problem.
The content of the invention
The invention provides a kind of single image to the fog method based on gradient field, it can effectively improve existing gradient field
Result figure is dark in defogging algorithm and the problems such as cross-color.
In order to solve the above problems, technical solution of the invention is:Single image to the fog method based on gradient field,
Characterized in that, mainly comprising the following steps:
(1) original fog-degraded image is gathered, the gradient and corresponding transmittance figure of fog-degraded image is calculated;
(2) multiple suitable scale-values are set, big, small three yardsticks of neutralization are generally selected, larger scale-value can be to the figure that degrades
Distant view region as in carries out contrast enhancing, and less yardstick can strengthen close shot region;
(3) corresponding scale-value is updated in enhancing function, then the gradient of degraded image carried out not using enhancing function
With the enhancing of yardstick;
(4) the gradient field function based on brightness is constructed, and is applied on multiple dimensioned, enhanced gradient image is entered
Row amendment operation so that enhanced result figure more meets human eye vision effect;
(5) weight is obtained by minimizing gradient mean square error progress gradient field reconstruction operation to the image after processing in step (4)
Build rear image;
(6) ratio between original Misty Image each passage is considered, color is constructed and recovers the factor, and it is melted with weighting
Hop algorithm is combined, using the Weighted Fusion formula with color recovery coefficient, after several reconstructions of gained in step (5)
Image is weighted mixing operation, the image after being merged;
(7) dynamic range compression is carried out to the image of gained in step (6), obtains final enhanced mist elimination image.
Wherein, the detailed process of step (4) the gradient field function of the construction based on brightness is:Consider first original
The monochrome information of image, relative gradient is converted to by absolute gradient, and such as following formula is represented:
Wherein, the monochrome information of I (x, y) representing input images,The gradient information of representing input images, andRepresent relative gradient.For identical absolute gradient value, the relative gradient value of brightness larger part is smaller.Then structure
The gradient field processing function based on brightness is made, such as following formula is represented:
Wherein,For average relative gradient value, max () computing ensures that relative gradient value is less and stretched, and
The original contrast of larger holding.Parameter alpha (1>α>0) threshold value of control gradient stretching, β (1>β>0) amplitude of restrained stretching.
When α value is bigger, represents more Grad and be stretched, β value gets over hour, the amplitude for representing corresponding gradient stretching is bigger.
Wherein, the detailed process of step (6) the Weighted Fusion operation is:First between each passage of consideration original image
Ratio, introduce color recover the factor, can be represented with following formula:
Wherein, i ∈ { r, g, b }, Ci(x, y) represents the color recovery factor of some Color Channel in RGB color space, represents defeated
Enter the weight shared by three Color Channels of image.Then using the Weighted Fusion strategy with the color recovery factor to each face
Chrominance channel gradient field reconstructed image RnHandled, can be represented with following formula:
Wherein, R'i(x, y) is the image after the final Weighted Fusion of each passage.The introducing of the color recovery factor, can be effective
Improve the color fidelity of result figure.
Wherein, the detailed process of step (7) described dynamic range compression is:Partial pixel in image R' after Weighted Fusion
The brightness value of point may exceed normal indication range, because gradient field enhancing is easy to low dynamic image being changed into high
Dynamic image.Therefore, in order that final enhanced image may be displayed in normal scope, it can be represented with following formula:
Wherein, max (R') and min (R') represent the minimum and maximum brightness value in image R' respectively, and R strengthens for final output
Mist elimination image afterwards.
Compared with existing single image defogging technology, the advantage of the invention is that:
(1) present invention is when to fog-degraded image grad enhancement, it is contemplated that the monochrome information of original image, constructs based on bright
The gradient field function of feature is spent, and is applied on multiple dimensioned, each enhanced image of yardstick Gradient is modified
Operation, so as to get enhancing effect more meet human eye vision effect;
(2) present invention is employed in last mixing operation recovers the Weighted Fusion algorithm of the factor to the figure after reconstruction with color
As being handled, the color recovery factor introduces the ratio considered between original color image each passage, can effectively carry
The color fidelity of high result images, it is to avoid the phenomenon of cross-color occur in result images;
(3) present invention is finally employing simple and effective Weighted Fusion method, with existing gradient field defogging and enhancing algorithm
Compare, reduce the number of times of gradient field reconstruction, improve the execution efficiency of algorithm.
Brief description of the drawings
Fig. 1:For the flow block schematic illustration of the present invention.
Embodiment
Further the present invention will be described in detail with specific embodiment below in conjunction with the accompanying drawings.
As shown in figure 1, the present invention is a kind of single image to the fog method based on gradient field, it is characterised in that main bag
Include following steps:
(1) original fog-degraded image is gathered, the gradient G and corresponding transmittance figure t of fog-degraded image is calculated;Herein
Gradient includes gradient horizontally and vertically, and the computational methods of transmittance figure are available to be expressed as below:
Wherein, p represents the coordinate of two-dimensional space, refers to the position of pixel in image, and max () computing represents to take maximum to transport
Calculate median () and represent median filtering operation, sv is the window size that medium filtering is used, and W (p)=min (I (p)) represents defeated
Enter minimum value of each pixel in three passages of image, t (p) represents the transmissivity of required greasy weather figure.
(2) three scale-values are selected, big, small three yardsticks of neutralization are represented respectively, larger scale-value can be to degraded image
In distant view region carry out contrast enhancing, and less yardstick can strengthen close shot region.
(3) corresponding scale-value is updated in enhancing function, then the gradient to fog-degraded image uses enhancing letter
Number carries out the enhancing of different scale, that is, calculates GC (t), wherein enhancing function C (t)=eλ(1-t)。
(4) consider the monochrome information of original image, absolute gradient is converted into relative gradient, such as following formula is represented:
Wherein, the monochrome information of I (x, y) representing input images,The gradient information of representing input images, andRepresent relative gradient.For identical absolute gradient value, the relative gradient value of brightness larger part is smaller.Then structure
The gradient field processing function based on brightness is made, such as following formula is represented:
Wherein,For average relative gradient value, max () computing ensures that relative gradient value is less and stretched, and
The original contrast of larger holding.Parameter alpha (1>α>0) threshold value of control gradient stretching, β (1>β>0) amplitude of restrained stretching.
When α value is bigger, represents more Grad and be stretched, β value gets over hour, the amplitude for representing corresponding gradient stretching is bigger.
For most of images, two parameters are set to α=0.1, β=0.8.
Then the processing function of the gradient field based on brightness of construction is applied on multiple dimensioned, different scale is strengthened
Image afterwards is modified operation, can be represented with following formula:
G'n(x, y)=Gn(x,y)·Φ(x,y)
Wherein, n is the label of yardstick, and three yardsticks are selected in experiment, i.e. n takes 1,2,3 respectively;Φ (x, y) is represented to gradient Gn
The stretch function of (x, y) operation, G'n(x, y) represents revised gradient fields.It should be noted that revised gradient fields G'n
(x, y) and former gradient fields GnThe direction of (x, y) is identical, is simply changed in size.
(5) image after processing in step (4) is obtained by minimizing gradient mean square error progress gradient field reconstruction operation
Image after to reconstruction, the process that gradient field is rebuild is represented by following process:
Assuming that known gradient G, H=f (x, y) are the image after its reconstruction.Gradient mean square error is minimized to be represented by:
Wherein,For image f (x, y) gradient, can with its first-order partial derivative come approximate representation, therefore,It can use
Following formula is represented:
Above formula need to meet following Euler-Lagrange equations:
F is substituted into can obtain in Euler-Lagrange equations:
Abbreviation is carried out to above formula again, poisson differential equation is obtained:
Wherein,Represent Laplace operator:Div represents divergence:
Just the image H after being rebuild by solving poisson differential equation.
For follow-up convenient expression, each Color Channel is rebuild to obtained image and is designated asWherein i represents Color Channel, n
Represent scale designation.
(6) ratio between original Misty Image each passage is considered, color is introduced and recovers the factor, following formula can be used
Represent:
Wherein, i ∈ { r, g, b }, Ci(x, y) represents the color recovery factor of some Color Channel in RGB color space, represents defeated
Enter the weight shared by three Color Channels of image.Then using the Weighted Fusion strategy with the color recovery factor to each face
Chrominance channel gradient field reconstruction imageHandled, can be represented with following formula:
Wherein, R'i(x, y) is the image after the final Weighted Fusion of each passage.The introducing of the color recovery factor, can be effective
Improve the color fidelity of result figure.
(7) because low dynamic image is easily converted to high-dynamics image by the image procossing of gradient field, it is therefore desirable to step
Suddenly the image of gained carries out dynamic range compression in (6), obtains the image after final defogging, can be represented with following formula:
Wherein, the minimum and maximum brightness value in max (R') and min (R') difference representative images R', R is that final output strengthens
Mist elimination image afterwards.
Claims (3)
1. a kind of single image to the fog method based on gradient field, it is characterised in that comprise the following steps:
(1) original fog-degraded image is gathered, the gradient G and corresponding transmittance figure t of fog-degraded image is calculated;
(2) multiple suitable scale-values are set, big, small three yardsticks of neutralization are generally selected, larger scale-value can drop to the greasy weather
Distant view region in matter image carries out contrast enhancing, and less yardstick can strengthen close shot region;
(3) corresponding scale-value is updated in enhancing function, then the gradient of degraded image carried out not using enhancing function
With the enhancing of yardstick;
(4) the gradient field function based on brightness is constructed, and is applied on multiple dimensioned, enhanced gradient image is entered
Row amendment operation so that enhanced result figure more meets human eye vision effect;
(5) weight is obtained by minimizing gradient mean square error progress gradient field reconstruction operation to the image after processing in step (4)
Build rear image;
(6) ratio between original Misty Image each passage is considered, color is constructed and recovers the factor, and it is melted with weighting
Hop algorithm is combined, using the Weighted Fusion formula with color recovery coefficient, after several reconstructions of gained in step (5)
Image is weighted mixing operation, the image after being merged;
(7) dynamic range compression is carried out to the image of gained in step (6), obtains the image after final defogging.
2. the single image to the fog method according to claim 1 based on gradient field, it is characterised in that in step (4),
Consider the monochrome information of original image, absolute gradient is replaced with relative gradient, and set lower limit to ensure that relative gradient value is smaller
Stretched, and the original contrast of larger holding, the gradient field function based on brightness is expressed as below:
<mfenced open = "{" close = "">
<mtable>
<mtr>
<mtd>
<mrow>
<mi>&Phi;</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mi>m</mi>
<mi>a</mi>
<mi>x</mi>
<mrow>
<mo>(</mo>
<mn>1</mn>
<mo>,</mo>
<msup>
<mrow>
<mo>(</mo>
<mfrac>
<mrow>
<mi>&alpha;</mi>
<mo>&CenterDot;</mo>
<mover>
<mrow>
<mo>|</mo>
<mo>|</mo>
<mo>&dtri;</mo>
<msup>
<mi>I</mi>
<mo>&prime;</mo>
</msup>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>|</mo>
<mo>|</mo>
</mrow>
<mo>&OverBar;</mo>
</mover>
</mrow>
<mrow>
<mo>|</mo>
<mo>|</mo>
<mo>&dtri;</mo>
<msup>
<mi>I</mi>
<mo>&prime;</mo>
</msup>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>|</mo>
<mo>|</mo>
</mrow>
</mfrac>
<mo>)</mo>
</mrow>
<mrow>
<mn>1</mn>
<mo>-</mo>
<mi>&beta;</mi>
</mrow>
</msup>
<mo>)</mo>
</mrow>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mrow>
<mo>|</mo>
<mo>|</mo>
<mo>&dtri;</mo>
<msup>
<mi>I</mi>
<mo>&prime;</mo>
</msup>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>|</mo>
<mo>|</mo>
<mo>=</mo>
<mfrac>
<mrow>
<mo>|</mo>
<mo>|</mo>
<mo>&dtri;</mo>
<mi>I</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>|</mo>
<mo>|</mo>
</mrow>
<mrow>
<mi>I</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
</mrow>
</mfrac>
</mrow>
</mtd>
</mtr>
</mtable>
</mfenced>
Wherein, I (x, y) represents the monochrome information of original fog-degraded image,The gradient information of representing input images,
AndRepresentative image relative gradient information,For average relative gradient value, max () computing ensures relatively terraced
Angle value is less to be stretched, and the original contrast of larger holding.Parameter alpha (1>α>0) threshold value of control gradient stretching, β
(1>β>0) amplitude of restrained stretching.When α value is bigger, represents more Grad and be stretched, β value is got over hour, represent correspondence
Gradient stretching amplitude it is bigger.
3. the single image to the fog method according to claim 1 based on gradient field, it is characterised in that in step (6),
Fusion is not weighted directly, but considers the ratio between original three passages of fog-degraded image, and construction color is extensive
Multifactor, is represented by:
<mrow>
<msub>
<mi>C</mi>
<mi>i</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfrac>
<mrow>
<msub>
<mi>I</mi>
<mi>i</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
</mrow>
<mrow>
<msub>
<mi>I</mi>
<mi>r</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>+</mo>
<msub>
<mi>I</mi>
<mi>g</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>+</mo>
<msub>
<mi>I</mi>
<mi>b</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
</mrow>
</mfrac>
</mrow>
Wherein, i ∈ { r, g, b }, Ci(x, y) represents the color recovery factor of some Color Channel in coloured image rgb space, generation
Weight shared by three Color Channels of table input picture.Then the color recovery factor is combined to ladder with Weighted Fusion algorithm
Image progress after degree domain is rebuild handles the image R after being mergedi' (x, y), it can be represented with following formula:
<mrow>
<msubsup>
<mi>R</mi>
<mi>i</mi>
<mo>&prime;</mo>
</msubsup>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<msub>
<mi>C</mi>
<mi>i</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>&CenterDot;</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>n</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>N</mi>
</munderover>
<msub>
<mi>w</mi>
<mi>n</mi>
</msub>
<msub>
<mi>R</mi>
<msub>
<mi>n</mi>
<mi>i</mi>
</msub>
</msub>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
</mrow>
1
Wherein, N is yardstick number, wnFor weight, RnImage after being rebuild for gradient field.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710301848.1A CN107146209A (en) | 2017-05-02 | 2017-05-02 | A kind of single image to the fog method based on gradient field |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710301848.1A CN107146209A (en) | 2017-05-02 | 2017-05-02 | A kind of single image to the fog method based on gradient field |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107146209A true CN107146209A (en) | 2017-09-08 |
Family
ID=59775404
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710301848.1A Pending CN107146209A (en) | 2017-05-02 | 2017-05-02 | A kind of single image to the fog method based on gradient field |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107146209A (en) |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108122210A (en) * | 2017-12-19 | 2018-06-05 | 长沙全度影像科技有限公司 | A kind of low-light (level) car plate video image enhancing method based on Retinex and enhancing gradient |
CN108648409A (en) * | 2018-04-28 | 2018-10-12 | 北京环境特性研究所 | A kind of smog detection method and device |
CN108734686A (en) * | 2018-05-28 | 2018-11-02 | 成都信息工程大学 | Multi-focus image fusing method based on Non-negative Matrix Factorization and visual perception |
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 |
CN111311502A (en) * | 2019-09-04 | 2020-06-19 | 中国科学院合肥物质科学研究院 | Method for processing foggy day image by using bidirectional weighted fusion |
CN111640079A (en) * | 2020-06-03 | 2020-09-08 | 徐州工程学院 | Defogging method based on image gradient distribution prior |
CN112734679A (en) * | 2021-01-26 | 2021-04-30 | 西安理工大学 | Fusion defogging method for medical operation video images |
CN113284061A (en) * | 2021-05-17 | 2021-08-20 | 大连海事大学 | Underwater image enhancement method based on gradient network |
CN115526808A (en) * | 2022-10-28 | 2022-12-27 | 东北林业大学 | Improved low-illumination image enhancement method of anisotropic diffusion equation model |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20140198992A1 (en) * | 2013-01-15 | 2014-07-17 | Apple Inc. | Linear Transform-Based Image Processing Techniques |
CN103942758A (en) * | 2014-04-04 | 2014-07-23 | 中国人民解放军国防科学技术大学 | Dark channel prior image dehazing method based on multiscale fusion |
CN104240192A (en) * | 2013-07-04 | 2014-12-24 | 西南科技大学 | Rapid single-image defogging algorithm |
-
2017
- 2017-05-02 CN CN201710301848.1A patent/CN107146209A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20140198992A1 (en) * | 2013-01-15 | 2014-07-17 | Apple Inc. | Linear Transform-Based Image Processing Techniques |
CN104240192A (en) * | 2013-07-04 | 2014-12-24 | 西南科技大学 | Rapid single-image defogging algorithm |
CN103942758A (en) * | 2014-04-04 | 2014-07-23 | 中国人民解放军国防科学技术大学 | Dark channel prior image dehazing method based on multiscale fusion |
Non-Patent Citations (3)
Title |
---|
DANIEL J. JOBSON: "A Multiscale Retinex for Bridging the Gap Between Color Images and the Human Observation of Scenes", 《IEEE TRANSACTIONS ON IMAGE PROCESSING》 * |
潘磊: "基于多尺度图像融合的单幅图像去雾算法", 《计算机工程》 * |
许欣: "结合视觉感知特性的梯度域图像增强方法", 《计算机辅助设计与图形学学报》 * |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108122210A (en) * | 2017-12-19 | 2018-06-05 | 长沙全度影像科技有限公司 | A kind of low-light (level) car plate video image enhancing method based on Retinex and enhancing gradient |
CN108648409A (en) * | 2018-04-28 | 2018-10-12 | 北京环境特性研究所 | A kind of smog detection method and device |
CN108648409B (en) * | 2018-04-28 | 2020-07-24 | 北京环境特性研究所 | Smoke detection method and device |
CN108734686A (en) * | 2018-05-28 | 2018-11-02 | 成都信息工程大学 | Multi-focus image fusing method based on Non-negative Matrix Factorization and visual perception |
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 |
CN111311502A (en) * | 2019-09-04 | 2020-06-19 | 中国科学院合肥物质科学研究院 | Method for processing foggy day image by using bidirectional weighted fusion |
CN111640079A (en) * | 2020-06-03 | 2020-09-08 | 徐州工程学院 | Defogging method based on image gradient distribution prior |
CN112734679A (en) * | 2021-01-26 | 2021-04-30 | 西安理工大学 | Fusion defogging method for medical operation video images |
CN113284061A (en) * | 2021-05-17 | 2021-08-20 | 大连海事大学 | Underwater image enhancement method based on gradient network |
CN113284061B (en) * | 2021-05-17 | 2024-04-05 | 大连海事大学 | Underwater image enhancement method based on gradient network |
CN115526808A (en) * | 2022-10-28 | 2022-12-27 | 东北林业大学 | Improved low-illumination image enhancement method of anisotropic diffusion equation model |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107146209A (en) | A kind of single image to the fog method based on gradient field | |
CN109741256B (en) | Image super-resolution reconstruction method based on sparse representation and deep learning | |
Jian et al. | Multi-scale image fusion through rolling guidance filter | |
CN103955905B (en) | Based on the single image to the fog method that fast wavelet transform and weighted image merge | |
CN102831591B (en) | Gaussian filter-based real-time defogging method for single image | |
CN108269244B (en) | Image defogging system based on deep learning and prior constraint | |
CN105205794B (en) | A kind of synchronous enhancing denoising method of low-light (level) image | |
CN102750674A (en) | Video image defogging method based on self-adapting allowance | |
CN103914820A (en) | Image haze removal method and system based on image layer enhancement | |
CN103049888A (en) | Image/video demisting method based on combination of dark primary color of atmospheric scattered light | |
CN102243758A (en) | Fog-degraded image restoration and fusion based image defogging method | |
CN105976337B (en) | A kind of image defogging method based on intermediate value guiding filtering | |
Liu et al. | Rank-one prior: Real-time scene recovery | |
Wang et al. | Variational single nighttime image haze removal with a gray haze-line prior | |
CN103996178A (en) | Sand and dust weather color image enhancing method | |
CN104574328A (en) | Color image enhancement method based on histogram segmentation | |
CN110634147A (en) | Image matting method based on bilateral boot up-sampling | |
CN106600547A (en) | Underwater image restoration method | |
CN108022225A (en) | Based on the improved dark channel prior image defogging algorithm of quick Steerable filter | |
CN104867121A (en) | Fast image defogging method based on dark channel prior and Retinex theory | |
CN106023108A (en) | Image defogging algorithm based on boundary constraint and context regularization | |
Kim et al. | Single image haze removal using hazy particle maps | |
CN111598814B (en) | Single image defogging method based on extreme scattering channel | |
CN103310486A (en) | Reconstruction method of atmospheric turbulence degraded images | |
CN111462022B (en) | Underwater image sharpness enhancement method |
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 |
Application publication date: 20170908 |
|
RJ01 | Rejection of invention patent application after publication |