CN104091310A - Image defogging method and device - Google Patents
Image defogging method and device Download PDFInfo
- Publication number
- CN104091310A CN104091310A CN201410286133.XA CN201410286133A CN104091310A CN 104091310 A CN104091310 A CN 104091310A CN 201410286133 A CN201410286133 A CN 201410286133A CN 104091310 A CN104091310 A CN 104091310A
- Authority
- CN
- China
- Prior art keywords
- image
- mist
- sampling
- transmissivity
- 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.)
- Pending
Links
Abstract
The invention provides an image defogging removing method and device. The method comprises the steps that a foggy image is down-sampled according to the size of the foggy image to obtain a down-sampling image; a transmittance graph and atmospheric light of the down-sampling image are estimated according to a fog model defogging method; the transmittance graph of the down-sampling image is up-sampled, and a transmittance graph of the foggy image is obtained; the fog model defogging method is adopted, and the atmospheric light and the transmittance graph of the foggy image are adopted to defog the foggy image. The image defogging method and device can defog images, and computing amount in processing is little.
Description
Technical field
The present invention relates to digital image processing techniques field, relate in particular to a kind of image defogging method capable and device.
Background technology
In haze sky situation, due to the low visibility of scene, the feature such as target contrast and color is attenuated, cause in the time of outdoor taking pictures, picture quality cannot meet user's needs, so need to process to eliminate the impact that weather brings to image, therefore image mist elimination is treated as the emphasis of people's research.
The current method for the processing of mist image roughly can be divided into two classes: a class is the method based on figure image intensifying, and these class methods are not considered the forming process that mist image is concrete, only chooses interested part in image and strengthens.Conventional image enchancing method has histogram equalization, homomorphic filtering and Retinex algorithm etc., these class methods are not considered the corresponding relation of Misty Image contrast and the scenery degree of depth, the enhancing effect of the image that scenic focal point thing change in depth is larger is undesirable, and the tone of image is changed, make not nature of image.
Defogging method capable based on mist model is mist image to be carried out to once contrary with imaging inverse process recover without mist image.Two transmissivities that primary unknowns is atmospheric parameter and image of mist model, the wherein degree of depth exponent function relation of transmissivity and image.Common defogging method capable is first atmospheric parameter and transmissivity to be estimated, then recovers without mist image according to imaging model.Compare the method for figure image intensifying, these class methods are with strong points, and the image ratio obtaining is more natural, and generally do not have information loss, can obtain good mist elimination effect, but this type of algorithm calculated amount is large at present, has limited the widespread use of this algorithm at engineering field.
Summary of the invention
The invention provides a kind of image mist elimination disposal route, can carry out mist elimination processing to image, and the calculated amount of processing is less.
The present invention also provides a kind of image mist elimination treating apparatus, can carry out mist elimination processing to image, and the calculated amount of processing is less.
Technical scheme of the present invention is achieved in that
A kind of image mist elimination disposal route, comprising:
According to having the size of mist image to there being mist image to carry out down-sampling, obtain down-sampled images;
Use mist model defogging method capable to estimate transmissivity figure and the atmosphere light of described down-sampled images;
The transmissivity figure of described down-sampled images is carried out to up-sampling, obtain the transmissivity figure of mist image;
Use mist model defogging method capable, and adopt described atmosphere light and have the transmissivity figure of mist image to have mist image to carry out mist elimination processing to described.
In said method, color space data form corresponding to down-sampled images is matrix Is; Wherein, the line number of Is equals the line number of pixel in down-sampled images, and the columns of Is equals the columns of pixel in down-sampled images, and each element in Is equals the value of respective pixel in down-sampled images;
Use mist model defogging method capable to estimate that the transmissivity figure of described down-sampled images and the step of atmosphere light can comprise:
A, matrix Ig corresponding to calculating gray level image; Account form is:
In the time having mist image to be gray level image, Ig=Is;
In the time having mist image to be RGB image, the line number of Ig equals the line number of Is, and the columns of Ig equals the columns of Is, and in Ig, each element equals the minimum value of R component, G component and the B component of corresponding pixel points in Is;
B, Ig is carried out to mean filter, obtain Ia, wherein, the radius of corresponding mean filter=(min (w, h)/400*7), wherein, the columns that w is Ig, the line number that h is Ig;
C, ask the average m of all elements in Ia;
Corresponding matrix L (x) and the atmosphere light A of transmissivity figure of D, calculating down-sampled images, account form is:
L (X)=min (min (ρ m, 0.9) Ia, Ig), wherein,
While being 2 matrixes in the bracket of min (), min () represents to ask a new matrix, and each element of this matrix is the minimum value of 2 matrix corresponding elements in bracket;
While being 1 matrix in the bracket of max (), max () represents to ask the maximal value of all elements in this matrix;
The account form of Im is:
In the time having mist image to be gray level image, Im=Is;
In the time having mist image to be RGB image, the line number of Im equals the line number of Is, and the columns of Im equals the columns of Is, and in Im, each element equals the maximal value of R component, G component and the B component of corresponding pixel points in Is.
Use mist model defogging method capable, and adopt described atmosphere light and have the transmissivity figure of mist image to have the mode that mist image carries out mist elimination processing to be to described:
In the time having mist image to be gray level image, calculate matrix corresponding to mist elimination image after treatment
obtain corresponding mist elimination image after treatment; Wherein,
I is the matrix that has mist image corresponding;
L' is matrix corresponding to transmissivity figure that has mist image;
X is that the line number line number that equals I, columns and all elements that columns equals I are 1 matrix;
In the time having mist image to be RGB image, calculate respectively R component, G component and the matrix corresponding to B component of the rear image of mist elimination processing
obtain corresponding mist elimination image after treatment; Wherein,
Ic is R component, G component or matrix corresponding to B component that has mist image;
L' is matrix corresponding to transmissivity figure that has mist image;
X is that the line number line number that equals I, columns and all elements that columns equals I are 1 matrix.
In said method, in the time having mist image to be gray level image, directly to there being mist image to carry out down-sampling;
In the time having mist image to be coloured image, will there is mist image to be converted to rgb color space data, then carry out down-sampling;
Down-sampling progression selects according to picture size or operational efficiency is selected;
Down-sampling mode is arest neighbors method of interpolation, bilinear interpolation or bicubic interpolation method.
Up-sampling progression is consistent with down-sampling progression; Up-sampling mode can be bilinear interpolation or bicubic interpolation method.
Said method may further include:
In the time having mist image to be gray level image, mist elimination image after treatment is carried out to brightness enhancing and contrast enhancing;
In the time that pending image is RGB image, image after treatment mist elimination is converted to yuv data form, Y component is carried out to brightness enhancing and contrast enhancing.
For the image after strengthening, the intensity of the mode estimating noise of input image by statistical picture flat site variance; Determine whether and carry out denoising by the intensity of picture noise of estimating; As need carry out denoising, adopt the mode of Steerable filter to carry out denoising, wherein the parameter of Steerable filter is set according to the intensity of described picture noise.
Image after denoising is converted into rgb format or jpeg format and preserves.
A kind of image mist elimination treating apparatus, comprising:
Down sample module, for according to having the size of mist image to there being mist image to carry out down-sampling, obtains down-sampled images;
Parameter estimation module, for being used mist model defogging method capable to estimate transmissivity figure and the atmosphere light of described down-sampled images;
Up-sampling module, for the transmissivity figure of described down-sampled images is carried out to up-sampling, obtains the transmissivity figure of mist image;
Mist elimination module, for using mist model defogging method capable, and adopts described atmosphere light and has the transmissivity figure of mist image to have mist image to carry out mist elimination processing to described.
In said apparatus, down sample module can be for:
In the time having mist image to be gray level image, directly to there being mist image to carry out down-sampling;
In the time having mist image to be coloured image, will there is mist image to be converted to rgb color space data, then carry out down-sampling;
Down-sampling progression selects according to picture size or operational efficiency is selected;
Down-sampling mode is arest neighbors method of interpolation, bilinear interpolation or bicubic interpolation method.
The Sampling series that up-sampling module adopts is consistent with down-sampling progression; Up-sampling mode can be bilinear interpolation or bicubic interpolation method.
Said apparatus may further include:
Strengthen module, in the time having mist image to be gray level image, mist elimination image after treatment is carried out to brightness enhancing and contrast enhancing; In the time that pending image is RGB image, image after treatment mist elimination is converted to yuv data form, Y component is carried out to brightness enhancing and contrast enhancing.
Denoising module, for the image for after strengthening, the intensity of the mode estimating noise of input image by statistical picture flat site variance; Determine whether and carry out denoising by the intensity of picture noise of estimating; As need carry out denoising, adopt the mode of Steerable filter to carry out denoising, wherein the parameter of Steerable filter is set according to the intensity of described picture noise.
Format converting module, for being converted into the image after denoising rgb format or jpeg format and preserving.
Visible, image mist elimination disposal route and device that the present invention proposes, by there being mist image to carry out down-sampling, and estimate transmissivity figure and the atmosphere light of down-sampled images, again the transmissivity figure of down-sampled images is carried out to up-sampling afterwards, obtain the transmissivity figure of mist image; And adopt and have the transmissivity figure of mist image and atmosphere light to there being mist image to process, can under the less prerequisite of calculated amount, realize the processing of image mist elimination.
Brief description of the drawings
Fig. 1 is the image mist elimination disposal route realization flow figure that the present invention proposes;
Fig. 2 is the process flow diagram of the embodiment of the present invention one;
Fig. 3 is the image mist elimination treating apparatus structural representation that the present invention proposes.
Embodiment
The present invention proposes a kind of image mist elimination disposal route, as the realization flow figure that Fig. 1 is the method, comprising:
Step 101: according to having the size of mist image to there being mist image to carry out down-sampling, obtain down-sampled images;
Step 102: use mist model defogging method capable to estimate transmissivity figure and the atmosphere light of described down-sampled images;
Step 103: the transmissivity figure of described down-sampled images is carried out to up-sampling, obtain the transmissivity figure of mist image;
Step 104: use mist model defogging method capable, and adopt described atmosphere light and have the transmissivity figure of mist image to have mist image to carry out mist elimination processing to described.
Afterwards, can further carry out brightness enhancing, contrast enhancing and denoising to the image after mist elimination; Image after finishing dealing with is the most at last converted into suitable form and preserves.
The original mist image that has can be gray level image or RGB image, has a mist image for different, and above-mentioned processing mode is also different.Below lifting specific embodiment introduces in detail.
Embodiment mono-:
As the realization flow figure that Fig. 2 is the present embodiment, the present embodiment comprises the following steps:
The first step, reads the original mist image that has, and is transformed to rgb color space data layout and obtains I;
Second step, according to the original size that has mist picture size, carries out one-level or secondary down-sampling by image I, obtains Is corresponding to down-sampled images.In the present embodiment, be less than 5,000,000 the mist image that has for picture traverse * picture altitude and adopt one-level down-sampling, the figure image width height after sampling is original half size; Picture traverse * picture altitude is greater than to 5,000,000 the mist image that has and adopts secondary down-sampling, the figure image width height after sampling is 1/4th original sizes, down-sampling progression can require to select according to Performance and quality, uses neighbor interpolation technology to carry out down-sampling in the present embodiment; Bilinear interpolation or the bicubic interpolation method of can also sampling carried out down-sampling.
Step 3, asks R, the G of each pixel in down-sampled images homography Is, the minimum value of tri-components of B, obtains the matrix that gray level image is corresponding, is designated as Ig; If Is is gray level image (having mist image is also gray level image), make Ig=Is.
Step 4, carries out mean filter to Ig and obtains Ia, the Size dependence of filtering window size and Ig, and filter radius can be set as (min (w, h)/400*7), wherein, the columns that w is Ig, the line number that h is Ig.
Step 5, asks for the average m of all elements in Ia.
Step 6, obtains the matrix L (x) corresponding to transmission plot of down-sampled images according to formula (1), wherein ρ be customized parameter and
in the present embodiment, ρ gets 1.3;
L(X)＝min(min(ρm,0.9)Ia,Ig) (1)
Wherein, while being 2 matrixes in the bracket of min (), min () represents to ask a new matrix, and each element of this matrix is the minimum value of 2 matrix corresponding elements in bracket;
Step 7, asks R, the G of each pixel in down-sampled images homography Is, the maximal value of tri-components of B, and obtain gray level image and be designated as Im, if Is is gray level image, Im=Is, obtains atmosphere light A according to formula (2):
Wherein, while being 1 matrix in the bracket of max (), max () represents to ask the maximal value of all elements in this matrix.
Step 8, the transmissivity figure of down-sampled images is carried out to up-sampling, obtain the matrix L that the final transmissivity figure that has mist image is corresponding ', in the present embodiment, use bilinear interpolation amplifying technique to carry out up-sampling, the multiple of up-sampling is consistent with the down-sampling multiple in second step; Can also adopt bicubic interpolation method to carry out up-sampling.
Step 9, the transmissivity figure that samples above-mentioned atmosphere light and have a mist image is to there being mist image to carry out mist elimination processing.Concrete mode is:
If having mist image is gray level image, adopt formula (3) to calculate matrix corresponding to mist elimination image after treatment, obtain corresponding mist elimination image after treatment;
Wherein, I is the matrix that has mist image corresponding;
L' is matrix corresponding to transmissivity figure that has mist image;
X is that the line number line number that equals I, columns and all elements that columns equals I are 1 matrix.
Be RGB image as figure has mist image, adopt respectively formula (4) to calculate R component, G component and the matrix corresponding to B component of the rear image of mist elimination processing, obtain corresponding mist elimination image after treatment;
Wherein, Ic is R component, G component or matrix corresponding to B component that has mist image;
L' is matrix corresponding to transmissivity figure that has mist image;
X is that the line number line number that equals I, columns and all elements that columns equals I are 1 matrix.
Step 10, carries out brightness enhancing and contrast enhancing to mist elimination image after treatment.Concrete mode is:
When original while having mist image to be gray level image, mist elimination image after treatment is directly carried out to brightness enhancing and contrast strengthens;
While having mist image to be RGB image, image after treatment mist elimination is transformed into the color space of YUV when original from rgb space, re-uses that formula (5) carries out brightness enhancing and contrast strengthens;
Wherein, L
dmax, L
wmaxset according to the value of ρ m with b;
Particularly, in the time that the value of ρ m is greater than 0.8, L
dmaxget 110, L
wmaxget 230, b gets 0.3;
When the value of ρ m is greater than 0.6 and while being less than 0.8, L
dmaxget 100, L
wmaxget 230, b gets 0.4;
In the time that ρ m gets other values, L
dmaxget 100, L
wmaxget 230, b gets 0.5;
Y is the Y component value of single pixel, and Y' is the Y component value that carries out brightness enhancing and the rear single pixel of contrast enhancing.
Step 11, carries out denoising to the image after strengthening.Concrete mode is:
First enhancing image is divided into the image block of 16*16, adds up the variance of each, calculate the Noise Variance Estimation of front 5 minimum variance mean values as image, by noise variance, noise-removed filtering parameter is set; If variance is less than 0.3, do not carry out filtering; If variance is greater than 0.3, use Steerable filter technology to carry out denoising to Y ' component, guiding image is Y ' component itself, Steerable filter radius is 10, regular coefficient is set according to the variance of picture noise, in the time that variance is less than 1, regular coefficient is 50, and variance is more than or equal to 1, and to be less than 2 o'clock regular coefficients be 80, and it is 100 that variance is more than or equal to 2 o'clock regular coefficients;
Step 12, is converted to yuv data output format and preserves.If be output as rgb format data, carry out the format conversion of YUV to RGB, then save data; If be output as jpeg format data, carry out YUV to jpeg format conversion, then save data.
Corresponding said method, the present invention also proposes a kind of image mist elimination treating apparatus, as the structural representation that Fig. 3 is this device, comprising:
Down sample module 301, for according to having the size of mist image to there being mist image to carry out down-sampling, obtains down-sampled images;
Parameter estimation module 302, for being used mist model defogging method capable to estimate transmissivity figure and the atmosphere light of described down-sampled images;
Up-sampling module 303, for the transmissivity figure of described down-sampled images is carried out to up-sampling, obtains the transmissivity figure of mist image;
Mist elimination module 304, for using mist model defogging method capable, and adopts described atmosphere light and has the transmissivity figure of mist image to have mist image to carry out mist elimination processing to described.
In said apparatus, down sample module 301 specifically can be for:
In the time having mist image to be gray level image, directly to there being mist image to carry out down-sampling;
In the time having mist image to be coloured image, will there is mist image to be converted to rgb color space data, then carry out down-sampling;
Down-sampling progression selects according to picture size or operational efficiency is selected;
Down-sampling mode is arest neighbors method of interpolation, bilinear interpolation or bicubic interpolation method.
The Sampling series that up-sampling module 303 adopts is consistent with down-sampling progression; Up-sampling mode is bilinear interpolation or bicubic interpolation method.
Said apparatus may further include:
Strengthen module 305, in the time having mist image to be gray level image, mist elimination image after treatment is carried out to brightness enhancing and contrast enhancing; In the time that pending image is RGB image, image after treatment mist elimination is converted to yuv data form, Y component is carried out to brightness enhancing and contrast enhancing.
Denoising module 306, for the image for after strengthening, the intensity of the mode estimating noise of input image by statistical picture flat site variance; Determine whether and carry out denoising by the intensity of picture noise of estimating; As need carry out denoising, adopt the mode of Steerable filter to carry out denoising, wherein the parameter of Steerable filter is set according to the intensity of described picture noise.
Format converting module 307, for being converted into the image after denoising rgb format or jpeg format and preserving.
As fully visible, image mist elimination disposal route and device that the present invention proposes, according to the size of picture size, carry out down-sampling to image, and this can promote the efficiency of mist elimination processing greatly, meets the mist elimination processing of ten million pixel scale image; Solve because estimated bias causes the dark problem of figure kine bias by mist elimination image being carried out to brightness enhancing; Strengthen by mist elimination image being carried out to contrast the image border texture problem not clearly that solved; Mist elimination image is carried out denoising and has been improved the subjective quality of image.The present invention has effectively avoided the problem that mist elimination performance is low and mist elimination figure kine bias is dark, image is clear not and noise is excessive, has promoted the subjective vision impression of mist elimination image, makes practical engineering project application become possibility.
The foregoing is only preferred embodiment of the present invention, in order to limit the present invention, within the spirit and principles in the present invention not all, any amendment of making, be equal to replacement, improvement etc., within all should being included in the scope of protection of the invention.
Claims (14)
1. an image mist elimination disposal route, is characterized in that, described method comprises:
According to having the size of mist image to there being mist image to carry out down-sampling, obtain down-sampled images;
Use mist model defogging method capable to estimate transmissivity figure and the atmosphere light of described down-sampled images;
The transmissivity figure of described down-sampled images is carried out to up-sampling, obtain the transmissivity figure of mist image;
Use mist model defogging method capable, and adopt described atmosphere light and have the transmissivity figure of mist image to have mist image to carry out mist elimination processing to described.
2. method according to claim 1, is characterized in that, color space data form corresponding to described down-sampled images is matrix Is; Wherein, the line number of Is equals the line number of pixel in down-sampled images, and the columns of Is equals the columns of pixel in down-sampled images, and each element in Is equals the value of respective pixel in down-sampled images;
Use mist model defogging method capable to estimate that the transmissivity figure of described down-sampled images and the step of atmosphere light comprise:
A, matrix Ig corresponding to calculating gray level image; Account form is:
In the time having mist image to be gray level image, Ig=Is;
In the time having mist image to be RGB image, the line number of Ig equals the line number of Is, and the columns of Ig equals the columns of Is, and in Ig, each element equals the minimum value of R component, G component and the B component of corresponding pixel points in Is;
B, Ig is carried out to mean filter, obtain Ia, wherein, the radius of corresponding mean filter=(min (w, h)/400*7), wherein, the columns that w is Ig, the line number that h is Ig;
C, ask the average m of all elements in Ia;
Corresponding matrix L (x) and the atmosphere light A of transmissivity figure of D, calculating down-sampled images, account form is:
L (X)=min (min (ρ m, 0.9) Ia, Ig), wherein,
While being 2 matrixes in the bracket of min (), min () represents to ask a new matrix, and each element of this matrix is the minimum value of 2 matrix corresponding elements in bracket;
While being 1 matrix in the bracket of max (), max () represents to ask the maximal value of all elements in this matrix;
The account form of Im is:
In the time having mist image to be gray level image, Im=Is;
In the time having mist image to be RGB image, the line number of Im equals the line number of Is, and the columns of Im equals the columns of Is, and in Im, each element equals the maximal value of R component, G component and the B component of corresponding pixel points in Is.
3. method according to claim 2, is characterized in that, described use mist model defogging method capable, and adopt described atmosphere light and have the transmissivity figure of mist image to have the mode that mist image carries out mist elimination processing to be to described:
In the time having mist image to be gray level image, calculate matrix corresponding to mist elimination image after treatment
obtain corresponding mist elimination image after treatment; Wherein,
I is the matrix that has mist image corresponding;
L' is matrix corresponding to transmissivity figure that has mist image;
X is that the line number line number that equals I, columns and all elements that columns equals I are 1 matrix;
In the time having mist image to be RGB image, calculate respectively R component, G component and the matrix corresponding to B component of the rear image of mist elimination processing
obtain corresponding mist elimination image after treatment; Wherein,
Ic is R component, G component or matrix corresponding to B component that has mist image;
L' is matrix corresponding to transmissivity figure that has mist image;
X is that the line number line number that equals I, columns and all elements that columns equals I are 1 matrix.
4. according to the method described in claim 1,2 or 3, it is characterized in that,
In the time having mist image to be gray level image, directly to there being mist image to carry out down-sampling;
In the time having mist image to be coloured image, will there is mist image to be converted to rgb color space data, then carry out down-sampling;
Down-sampling progression selects according to picture size or operational efficiency is selected;
Down-sampling mode is arest neighbors method of interpolation, bilinear interpolation or bicubic interpolation method.
5. method according to claim 4, is characterized in that, described up-sampling progression is consistent with down-sampling progression; Up-sampling mode is bilinear interpolation or bicubic interpolation method.
6. according to the method described in claim 1,2 or 3, it is characterized in that, described method further comprises:
In the time having mist image to be gray level image, mist elimination image after treatment is carried out to brightness enhancing and contrast enhancing;
In the time that pending image is RGB image, image after treatment mist elimination is converted to yuv data form, Y component is carried out to brightness enhancing and contrast enhancing.
7. method according to claim 6, is characterized in that, described method further comprises:
For the image after strengthening, the intensity of the mode estimating noise of input image by statistical picture flat site variance; Determine whether and carry out denoising by the intensity of picture noise of estimating; As need carry out denoising, adopt the mode of Steerable filter to carry out denoising, wherein the parameter of Steerable filter is set according to the intensity of described picture noise.
8. method according to claim 7, is characterized in that, described method further comprises: the image after denoising is converted into rgb format or jpeg format and preserves.
9. an image mist elimination treating apparatus, is characterized in that, described device comprises:
Down sample module, for according to having the size of mist image to there being mist image to carry out down-sampling, obtains down-sampled images;
Parameter estimation module, for being used mist model defogging method capable to estimate transmissivity figure and the atmosphere light of described down-sampled images;
Up-sampling module, for the transmissivity figure of described down-sampled images is carried out to up-sampling, obtains the transmissivity figure of mist image;
Mist elimination module, for using mist model defogging method capable, and adopts described atmosphere light and has the transmissivity figure of mist image to have mist image to carry out mist elimination processing to described.
10. device according to claim 9, is characterized in that, described down sample module is used for:
In the time having mist image to be gray level image, directly to there being mist image to carry out down-sampling;
In the time having mist image to be coloured image, will there is mist image to be converted to rgb color space data, then carry out down-sampling;
Down-sampling progression selects according to picture size or operational efficiency is selected;
Down-sampling mode is arest neighbors method of interpolation, bilinear interpolation or bicubic interpolation method.
11. devices according to claim 10, is characterized in that, the Sampling series that described up-sampling module adopts is consistent with down-sampling progression; Up-sampling mode is bilinear interpolation or bicubic interpolation method.
12. according to the device described in claim 9,10 or 11, it is characterized in that, described device further comprises:
Strengthen module, in the time having mist image to be gray level image, mist elimination image after treatment is carried out to brightness enhancing and contrast enhancing; In the time that pending image is RGB image, image after treatment mist elimination is converted to yuv data form, Y component is carried out to brightness enhancing and contrast enhancing.
13. devices according to claim 12, is characterized in that, described device further comprises:
Denoising module, for the image for after strengthening, the intensity of the mode estimating noise of input image by statistical picture flat site variance; Determine whether and carry out denoising by the intensity of picture noise of estimating; As need carry out denoising, adopt the mode of Steerable filter to carry out denoising, wherein the parameter of Steerable filter is set according to the intensity of described picture noise.
14. devices according to claim 13, is characterized in that, described device further comprises:
Format converting module, for being converted into the image after denoising rgb format or jpeg format and preserving.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410286133.XA CN104091310A (en) | 2014-06-24 | 2014-06-24 | Image defogging method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410286133.XA CN104091310A (en) | 2014-06-24 | 2014-06-24 | Image defogging method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104091310A true CN104091310A (en) | 2014-10-08 |
Family
ID=51639025
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410286133.XA Pending CN104091310A (en) | 2014-06-24 | 2014-06-24 | Image defogging method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104091310A (en) |
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104767912A (en) * | 2015-04-14 | 2015-07-08 | 深圳市晶日盛科技有限公司 | Video defogging method and system based on FPGA |
| CN104796582A (en) * | 2015-04-20 | 2015-07-22 | 中国矿业大学 | Video image denoising and enhancing method and device based on random ejection retinex |
| CN105898111A (en) * | 2016-05-06 | 2016-08-24 | 西安理工大学 | Video defogging method based on spectral clustering |
| CN106097259A (en) * | 2016-05-27 | 2016-11-09 | 安徽超远信息技术有限公司 | A kind of Misty Image fast reconstructing method based on absorbance optimisation technique |
| CN106251296A (en) * | 2016-07-06 | 2016-12-21 | 北京元心科技有限公司 | A kind of image defogging method and system |
| CN106780356A (en) * | 2016-11-15 | 2017-05-31 | 天津大学 | Image defogging method based on convolutional neural networks and prior information |
| CN107481199A (en) * | 2017-07-27 | 2017-12-15 | 广东欧珀移动通信有限公司 | Image defogging processing method, device, storage medium and mobile terminal |
| CN107610058A (en) * | 2017-08-28 | 2018-01-19 | 浙江工业大学 | High-definition picture defogging method based on down-sampling |
| CN107705262A (en) * | 2017-10-10 | 2018-02-16 | 中山大学 | A kind of defogging method based on mixing priori learning model |
| CN107730472A (en) * | 2017-11-03 | 2018-02-23 | 昆明理工大学 | A kind of image defogging optimized algorithm based on dark primary priori |
| CN108492259A (en) * | 2017-02-06 | 2018-09-04 | 联发科技股份有限公司 | A kind of image processing method and image processing system |
| CN108734670A (en) * | 2017-04-20 | 2018-11-02 | 天津工业大学 | The restoration algorithm of single width night weak illumination haze image |
| CN109345485A (en) * | 2018-10-22 | 2019-02-15 | 北京达佳互联信息技术有限公司 | A kind of image enchancing method, device, electronic equipment and storage medium |
| CN110544216A (en) * | 2019-08-29 | 2019-12-06 | 北京理工大学 | Video defogging system based on deep learning |
| CN111861936A (en) * | 2020-07-29 | 2020-10-30 | 北京字节跳动网络技术有限公司 | Image defogging method and device, electronic equipment and computer readable storage medium |
| CN112070701A (en) * | 2020-09-08 | 2020-12-11 | 北京字节跳动网络技术有限公司 | Image generation method, device, equipment and computer readable medium |
| CN112215789A (en) * | 2020-10-12 | 2021-01-12 | 北京字节跳动网络技术有限公司 | Image defogging method, device, equipment and computer readable medium |
| WO2021051239A1 (en) * | 2019-09-16 | 2021-03-25 | Covidien Lp | Systems and methods for real-time de-hazing in images |
| CN115412669A (en) * | 2022-08-26 | 2022-11-29 | 清华大学 | Foggy day imaging method and device based on image signal-to-noise ratio analysis |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102289791A (en) * | 2011-06-29 | 2011-12-21 | 清华大学 | Method for quickly demisting single image |
| WO2013029337A1 (en) * | 2011-08-30 | 2013-03-07 | Fujitsu Limited | Image defogging method and system |
| CN103279931A (en) * | 2013-06-03 | 2013-09-04 | 中国人民解放军国防科学技术大学 | Defogged image denoising method based on transmissivity |
| CN103824258A (en) * | 2012-11-19 | 2014-05-28 | 株式会社Pfu | Image processing apparatus and image processing method |
-
2014
- 2014-06-24 CN CN201410286133.XA patent/CN104091310A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102289791A (en) * | 2011-06-29 | 2011-12-21 | 清华大学 | Method for quickly demisting single image |
| WO2013029337A1 (en) * | 2011-08-30 | 2013-03-07 | Fujitsu Limited | Image defogging method and system |
| CN103824258A (en) * | 2012-11-19 | 2014-05-28 | 株式会社Pfu | Image processing apparatus and image processing method |
| CN103279931A (en) * | 2013-06-03 | 2013-09-04 | 中国人民解放军国防科学技术大学 | Defogged image denoising method based on transmissivity |
Non-Patent Citations (2)
| Title |
|---|
| 刘倩等: "基于单幅图像的快速去雾算法", 《第25届中国控制与决策会议论文集》 * |
| 章郡锋等: "基于暗原色先验去雾的改进算法", 《电视技术》 * |
Cited By (28)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104767912B (en) * | 2015-04-14 | 2018-11-30 | 深圳日盛科技股份有限公司 | Video defogging method and system based on FPGA |
| CN104767912A (en) * | 2015-04-14 | 2015-07-08 | 深圳市晶日盛科技有限公司 | Video defogging method and system based on FPGA |
| CN104796582A (en) * | 2015-04-20 | 2015-07-22 | 中国矿业大学 | Video image denoising and enhancing method and device based on random ejection retinex |
| CN104796582B (en) * | 2015-04-20 | 2018-05-04 | 中国矿业大学 | Video image denoising and Enhancement Method and device based on random injection retinex |
| CN105898111A (en) * | 2016-05-06 | 2016-08-24 | 西安理工大学 | Video defogging method based on spectral clustering |
| CN105898111B (en) * | 2016-05-06 | 2018-11-27 | 西安理工大学 | A kind of video defogging method based on spectral clustering |
| CN106097259A (en) * | 2016-05-27 | 2016-11-09 | 安徽超远信息技术有限公司 | A kind of Misty Image fast reconstructing method based on absorbance optimisation technique |
| CN106097259B (en) * | 2016-05-27 | 2018-11-30 | 安徽超远信息技术有限公司 | A kind of Misty Image fast reconstructing method based on transmissivity optimisation technique |
| CN106251296A (en) * | 2016-07-06 | 2016-12-21 | 北京元心科技有限公司 | A kind of image defogging method and system |
| CN106780356B (en) * | 2016-11-15 | 2020-04-28 | 天津大学 | Image defogging method based on convolutional neural network and prior information |
| CN106780356A (en) * | 2016-11-15 | 2017-05-31 | 天津大学 | Image defogging method based on convolutional neural networks and prior information |
| CN108492259A (en) * | 2017-02-06 | 2018-09-04 | 联发科技股份有限公司 | A kind of image processing method and image processing system |
| CN108734670B (en) * | 2017-04-20 | 2021-05-18 | 天津工业大学 | Method for restoring single night weak-illumination haze image |
| CN108734670A (en) * | 2017-04-20 | 2018-11-02 | 天津工业大学 | The restoration algorithm of single width night weak illumination haze image |
| CN107481199A (en) * | 2017-07-27 | 2017-12-15 | 广东欧珀移动通信有限公司 | Image defogging processing method, device, storage medium and mobile terminal |
| CN107610058A (en) * | 2017-08-28 | 2018-01-19 | 浙江工业大学 | High-definition picture defogging method based on down-sampling |
| CN107705262A (en) * | 2017-10-10 | 2018-02-16 | 中山大学 | A kind of defogging method based on mixing priori learning model |
| CN107730472A (en) * | 2017-11-03 | 2018-02-23 | 昆明理工大学 | A kind of image defogging optimized algorithm based on dark primary priori |
| CN109345485A (en) * | 2018-10-22 | 2019-02-15 | 北京达佳互联信息技术有限公司 | A kind of image enchancing method, device, electronic equipment and storage medium |
| CN109345485B (en) * | 2018-10-22 | 2021-04-16 | 北京达佳互联信息技术有限公司 | Image enhancement method and device, electronic equipment and storage medium |
| CN110544216A (en) * | 2019-08-29 | 2019-12-06 | 北京理工大学 | Video defogging system based on deep learning |
| WO2021051239A1 (en) * | 2019-09-16 | 2021-03-25 | Covidien Lp | Systems and methods for real-time de-hazing in images |
| CN111861936A (en) * | 2020-07-29 | 2020-10-30 | 北京字节跳动网络技术有限公司 | Image defogging method and device, electronic equipment and computer readable storage medium |
| CN111861936B (en) * | 2020-07-29 | 2023-03-24 | 抖音视界有限公司 | Image defogging method and device, electronic equipment and computer readable storage medium |
| CN112070701A (en) * | 2020-09-08 | 2020-12-11 | 北京字节跳动网络技术有限公司 | Image generation method, device, equipment and computer readable medium |
| CN112215789A (en) * | 2020-10-12 | 2021-01-12 | 北京字节跳动网络技术有限公司 | Image defogging method, device, equipment and computer readable medium |
| CN115412669A (en) * | 2022-08-26 | 2022-11-29 | 清华大学 | Foggy day imaging method and device based on image signal-to-noise ratio analysis |
| CN115412669B (en) * | 2022-08-26 | 2023-06-06 | 清华大学 | Foggy day imaging method and device based on image signal-to-noise ratio analysis |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104091310A (en) | Image defogging method and device | |
| CN110148095B (en) | Underwater image enhancement method and enhancement device | |
| WO2016206087A1 (en) | Low-illumination image processing method and device | |
| Li et al. | Single image dehazing using the change of detail prior | |
| CN101783012B (en) | Automatic image defogging method based on dark primary colour | |
| Ancuti et al. | Single image dehazing by multi-scale fusion | |
| US10672112B2 (en) | Method and system for real-time noise removal and image enhancement of high-dynamic range images | |
| EP2187620B1 (en) | Digital image processing and enhancing system and method with function of removing noise | |
| Pang et al. | Improved single image dehazing using guided filter | |
| US7983511B1 (en) | Methods and apparatus for noise reduction in digital images | |
| Ma et al. | An effective fusion defogging approach for single sea fog image | |
| CN106530257A (en) | Remote sensing image de-fogging method based on dark channel prior model | |
| CN105046658A (en) | Low-illumination image processing method and device | |
| CN110544213A (en) | Image defogging method based on global and local feature fusion | |
| CN107424133A (en) | Image defogging method, device, computer can storage medium and mobile terminals | |
| CN101877123A (en) | Image enhancement method and device | |
| CN112561804A (en) | Low-illumination underwater image enhancement method based on multi-scale detail enhancement | |
| US8488899B2 (en) | Image processing apparatus, method and recording medium | |
| CN111192205A (en) | Image defogging method and system and computer readable storage medium | |
| CN111462022B (en) | Underwater image sharpness enhancement method | |
| CN108109113A (en) | Single image to the fog method and device based on bilateral filtering and medium filtering | |
| CN103778605A (en) | Greasy weather image enhancement method | |
| CN103295205A (en) | Low-light-level image quick enhancement method and device based on Retinex | |
| Fuh et al. | Mcpa: A fast single image haze removal method based on the minimum channel and patchless approach | |
| Negru et al. | Exponential image enhancement in daytime fog conditions |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20141008 |
|
| WD01 | Invention patent application deemed withdrawn after publication |