CN103049888A - Image/video demisting method based on combination of dark primary color of atmospheric scattered light - Google Patents

Abstract

The invention discloses an image/video demisting method based on the combination of a dark primary color and atmospheric scattered light. The method comprises the following specific steps of: 1, inputting an original misty image I into an image processing system of a computer, and acquiring a dark primary color image Idark of the original misty image I; 2, according to the obtained dark primary color image Idark, evaluating the atmospheric light values A, i.e. airR, airG and airB of red, green and blue (RGB) channels of the original image; 3, evaluating the atmospheric scattered light value V (x,y) of the original misty image I; and 4, evaluating a demisted restored image J according to images of the RGB channels of the original misty image I, the atmospheric light values A and the atmospheric scattered light value V to obtain a demisted restored image J finally. According to the image/video demisting method based on the combination of the dark primary color and atmospheric scattered light, a clear demisted image is obtained by calculating the dark primary color image of the original image and restoring the original image with the atmospheric light values and the atmospheric scattered light value.

Description

Based on the image/video defogging method capable of dark primary in conjunction with atmospheric scattering light

Technical field

The invention belongs to the image processing and strengthen technical field, relate to a kind of based on the image/video defogging method capable of dark primary in conjunction with atmospheric scattering light.

Background technology

It is a very significant problem that the mist image sharpening is arranged, and under the condition of the low irradiances such as greasy weather, overcast and rainy and evening, the picture contrast of collection is very low, poor visibility.As: these classes such as intelligent traffic monitoring, topographic(al) reconnaissance need to monitor the situation of visual field, when in the greasy weather situation, because the low visibility of scene, target contrast and color characteristic are attenuated in the image, cause system to work, therefore need in image, eliminate fog to the impact of scene image.

At present, main image mist elimination disposal route can be divided into following two classes:

The first kind is that normal image strengthens algorithm, the figure image intensifying is divided into the color of image enhancing and picture contrast strengthens, and color of image strengthens main by color constancy algorithm and tone-mapping algorithm, for example: the brightness of image curve adjustment, the brightness of image linear stretch, histogram equalization and gamma algorithm; Algorithm for image enhancement mainly contains the frequency field Image Sharpening Algorithm, based on Image Sharpening Algorithm of mask etc.This class algorithm does not consider that the greasy weather atmosphere is on the impact of image.

Equations of The Second Kind is based on the method for atmospheric degradation physical model, this method need to obtain extraneous information, for example: the method that has need to utilize special-purpose proven radar installations to obtain depth information, then utilize view data and depth information to ask the parameter of physical model, then parameter is brought into degradation model, just can obtain estimated image; Some methods need to obtain the image of Same Scene under two kinds of different weather, could obtain depth information, and these require bad realization.

Existing mist elimination technology, some has specific requirement to input picture, some method requires the user to carry out alternately, this processes at realtime graphic and also is difficult in the application satisfy, some image adopts the image of single frames as input, but has the problem of recovering rear image generation cross-color, does not meet the requirement of mist elimination, the processing speed that also has is too slow, can't be applied in the real-time system.

Summary of the invention

The object of the present invention is to provide a kind of based on the image/video defogging method capable of dark primary in conjunction with atmospheric scattering light, by calculating the dark primary image of original image, utilize atmosphere light value and atmospheric scattering light value, recuperating original image obtains comparatively clearly image behind the mist elimination.

The technical solution adopted in the present invention is based on the image/video defogging method capable of dark primary in conjunction with atmospheric scattering light, specifically to implement according to following steps:

Step 1, with in the original image processing system that mist image I input computing machine arranged, ask for the original dark primary image I that the mist image I is arranged ^Dark

The dark primary image I that step 2, basis are obtained ^Dark, ask for the atmosphere light value A of three Color Channels of original image RGB, be respectively airR, airG and airB;

Step 3, ask for original atmospheric scattering light value V(x, the y that the mist image I is arranged);

Step 4, ask for restored image J behind the mist elimination according to original three channel image of RGB, atmosphere light value A and the atmospheric scattering light value V that the mist image I arranged, finally obtain the restored image J behind the mist elimination.

Characteristics of the present invention also are,

Step 1 is specifically implemented according to following steps:

Step 1.1, with in the original image processing system that mist image I input computing machine arranged, it is original that the pixel of mist image is arranged is I(x, y), there is the mist image I to separate with original, extraction obtains the original image that three Color Channels of RGB of mist image I are arranged, and is respectively: image I ^R, image I ^G, image I ^B

Step 1.2, the original image I that three Color Channels of mist image I are arranged to obtaining through step 1.1 ^R, image I ^G, image I ^BCarry out respectively mini-value filtering, namely obtain the image of filtered three Color Channels, be respectively image DR, image DG and image DB;

Filtered three images in step 1.3, the comparison step 1.2: i.e. image DR, image DG and image DB, choose the minimum value of three image corresponding pixel points as the pixel value I of dark primary image corresponding point ^Dark(x, y) namely obtains the original dark primary image I that the mist image I is arranged ^Dark, the original dark primary image I that the mist image I is arranged ^DarkThe value of each pixel is implemented by following algorithm:

$I^{\mathrm{dark}}(x,y)=\underset{c\∈\{R,G,B\}}{\min }\left(\underset{(x,y)\∈\mathrm{\Ω}(x^{\′},y^{\′})}{\min }\left(I^c(x^{\′},y^{\′})\right)\right),$

Wherein, I ^cFor the original Color Channel that the mist image I is arranged, be I ^R, I ^GAnd I ^B, Ω (x ', y ') be the zone of mini-value filtering.

Original the mist image I is arranged is coloured image, directly uses the original RGB three-component that the mist image is arranged, and need not to carry out the conversion of color space.

It is original that the mist image I is arranged is the picture that existed in the computing machine or the real time video data of video file or camera collection.

Step 2 is specifically implemented according to following steps:

Step 2.1, the original dark primary image I that the mist image I is arranged of obtaining according to step 1.3 ^DarkSeek out the histogram of dark primary image;

Step 2.2, according to the histogram of the dark primary image of step 2.1, in brightness value is front 0.1% dark primary image, seek the pixel value of brightness maximum as threshold value, behind the definite threshold in the histogram of dark primary image the selected pixels value greater than the pixel of threshold value;

Step 2.3, with the pixel chosen in the step 2.2 corresponding to the original RGB three-component image that the mist image I is arranged that obtains in the step 1.1, extract respectively the maximal value in the pixel value in the corresponding original RGB three-component image that the mist image I arranged of pixel, namely obtain atmosphere light value A, be respectively airR, airG and airB:

Step 3 is specifically implemented according to following steps:

Step 3.1, with the original image that three Color Channels of RGB of mist image I are arranged that obtains in the step 1.1, i.e. image I ^R, image I ^G, image I ^BIn the minimum value of each pixel respective pixel value extract:

That is: W(x, y)=min{I(x, y);

Step 3.2, to pixel value W(x, the y of the pixel that extracts in the step 3.1) carry out medium filtering, specifically implement according to following algorithm:

C（x，y）=median _sv（W（x，y））

Wherein, C(x, y) be the result of medium filtering, W(x, y) result that obtains for step 3.1, sv is the square window size of using in the median filter;

Step 3.3, according to the result that step 3.1 and step 3.2 obtain, in conjunction with atmospheric scattering light algorithm, namely obtain original atmospheric scattering light value V(x, the y that the mist image I is arranged), atmospheric scattering light is specifically implemented according to following algorithm:

B（x，y）=C-median _sv（︱W-C︱）（x，y）

V（x，y）=max（min（ρB（x，y），W（x，y）），0）

Wherein, B(x, y) expression W(x, y) local mean value and the difference of local standard deviation, ρ is for taking advantage of sex factor, the intensity of expression recovery, span are between 0.75 to 0.95, sv is that the square window of using in the median filter is big or small, value is 41.

Step 4 is specifically implemented according to following steps:

Step 4.1, with the original image that three Color Channels of RGB of mist image I are arranged in the step 1.1: image I ^R, image I ^GAnd image I ^BThe atmosphere light value A of the separately Color Channel that obtains in the integrating step 2 respectively, atmospheric scattering light value V(x, the y that step 3 obtains), use Image Restoration Algorithm, ask for obtaining image J(x, y behind the mist elimination) the image J of three Color Channels of RGB ^c(x, y), the algorithm of the restored image behind the mist elimination is as follows:

$J^c(x,y)=\frac{I^c(x,y)-V(x,y)}{1-V(x,y)/A},$

Wherein, J ^c(x, y) is the pixel value of the Color Channel correspondence image of image behind the mist elimination, I ^c(x, y) is the original pixel value that the Color Channel correspondence image of mist image I is arranged, and A is the atmosphere light value, V(x, y) be the atmospheric scattering light value;

Step 4.2, with the image of three Color Channels of RGB of the mist elimination image that obtains in the step 4.1, i.e. image J ^R, image J ^GWith image J ^BMake up the image J after namely obtaining restoring.

Beneficial effect of the present invention is,

(1) the inventive method is by calculating the original dark primary image that the mist image is arranged, and utilizes atmosphere light value and atmospheric scattering light value, and recuperating original image namely obtains comparatively clearly image behind the mist elimination.

(2) method of the present invention does not only need artificial participation, can also reduce significantly calculation cost, has saved computing time, when obtaining clearly visual effect, improves significantly image sharpening speed.

(3) can be widely used in the safe driving assistant system of video monitoring, topographic(al) reconnaissance and existing vehicle, aircraft, ship, can be applied in the higher system of some requirement of real-times.

Description of drawings

Fig. 1 is of the present invention based on the process flow diagram of dark primary in conjunction with the image/video defogging method capable of atmospheric scattering light.

Embodiment

The present invention is described in detail below in conjunction with the drawings and specific embodiments.

Of the present invention based on the image/video defogging method capable of dark primary in conjunction with atmospheric scattering light, to the original treatment scheme that the mist image arranged as shown in Figure 1, specifically implement according to following steps:

Step 1, with in the original image processing system that mist image I input computing machine arranged, ask for the original dark primary image I that the mist image I is arranged ^Dark:

Step 1.1, with in the original image processing system that mist image I input computing machine arranged, it is original that the pixel of mist image is arranged is I(x, y), there is the mist image I to separate with original, extraction obtains the original image that three Color Channels of RGB of mist image I are arranged, and is respectively: image I ^R, image I ^G, image I ^B

Step 1.2, the original image I that three Color Channels of mist image I are arranged to obtaining through step 1.1 ^R, image I ^G, image I ^BCarry out respectively mini-value filtering, namely obtain the image of filtered three Color Channels, be respectively image DR, image DG and image DB;

Filtered three images in step 1.3, the comparison step 1.2: i.e. image DR, image DG and image DB, choose the minimum value of three image corresponding pixel points as the pixel value I of dark primary image corresponding point ^Dark(x, y) namely obtains the original dark primary image I that the mist image I is arranged ^Dark, the original dark primary image I that the mist image I is arranged ^DarkThe value of each pixel is implemented by following algorithm:

$I^{\mathrm{dark}}(x,y)=\underset{c\∈\{R,G,B\}}{\min }\left(\underset{(x,y)\∈\mathrm{\Ω}(x^{\′},y^{\′})}{\min }\left(I^c(x^{\′},y^{\′})\right)\right),$

Wherein, I ^cFor the original Color Channel that the mist image I is arranged, be I ^R, I ^GAnd I ^B, Ω (x ', y ') be the zone of mini-value filtering.Originally can directly use the original RGB three-component that the mist image is arranged when having the mist image I to be coloured image, not need to carry out the conversion of color space.Wherein input original the mist image I is arranged both can be picture or the video file that has existed in the computing machine, also can be the real time video data of camera collection.

The dark primary image I that step 2, basis are obtained ^Dark, ask for the original atmosphere light value A that three Color Channels of RGB of mist image I are arranged, be respectively airR, airG and airB:

Step 2.1, the original dark primary image I that the mist image I is arranged of obtaining according to step 1.3 ^DarkSeek out the histogram of dark primary image;

Step 2.2, according to the histogram of the dark primary image of step 2.1, in brightness value is front 0.1% dark primary image, seek the pixel value of brightness maximum as threshold value, behind the definite threshold in the histogram of dark primary image the selected pixels value greater than the pixel of threshold value;

Step 2.3, with the pixel chosen in the step 2.2 corresponding to the original RGB three-component image that the mist image I is arranged that obtains in the step 1.1, extract respectively the maximal value in the pixel value in the corresponding original RGB three-component image that the mist image I arranged of pixel, namely obtain atmosphere light value A, be respectively airR, airG and airB.

Step 3, ask for original atmospheric scattering light value V(x, the y that the mist image I is arranged):

Step 3.1, with the original image that three Color Channels of RGB of mist image I are arranged that obtains in the step 1.1, i.e. image I ^R, image I ^G, image I ^BIn the minimum value of each pixel respective pixel value extract:

That is: W(x, y)=min{I(x, y);

Step 3.2, to pixel value W(x, the y of the pixel that extracts in the step 3.1) carry out medium filtering, specifically implement according to following algorithm:

C（x，y）=median _sv（W（x，y））

Wherein, C(x, y) be the result of medium filtering, W(x, y) result that obtains for step 3.1, sv is the square window size of using in the median filter;

Step 3.3, according to the result who obtains in step 3.1 and the step 3.2, in conjunction with atmospheric scattering light algorithm, namely obtain original atmospheric scattering light value V(x, the y that the mist image I is arranged), atmospheric scattering light is specifically implemented according to following algorithm:

B（x，y）=C-median _sv（︱W-C︱）（x，y）

V（x，y）=max（min（ρB（x，y），W（x，y）），0）

Wherein, B(x, y) expression W(x, y) local mean value and the difference of local standard deviation, ρ is for taking advantage of sex factor, the intensity of expression recovery, span are between 0.75 to 0.95, sv is that the square window of using in the median filter is big or small, value is 41.

Step 4, ask for restored image J behind the mist elimination according to original three channel image of RGB, atmosphere light value A and the atmospheric scattering light value V that the mist image I arranged, finally obtain the restored image J behind the mist elimination:

Obtain restored image and will restore respectively original three Color Channels of RGB that the mist image I is arranged, wherein atmosphere light value A is that three values are airR, airG and airB, and original have the image of three Color Channels of RGB of mist image I to be respectively: image I ^R, image I ^GAnd image I ^B

Step 4 is specifically implemented according to following steps:

Step 4.1, with the original image that three Color Channels of RGB of mist image I are arranged in the step 1.1: image I ^R, image I ^GAnd image I ^BThe atmosphere light value A of the separately Color Channel that obtains in the integrating step 2 respectively, atmospheric scattering light value V(x, the y that step 3 obtains), use Image Restoration Algorithm, ask for obtaining image J(x, y behind the mist elimination) the image J of three Color Channels of RGB ^c(x, y), the algorithm of the restored image behind the mist elimination is as follows:

$J^c(x,y)=\frac{I^c(x,y)-V(x,y)}{1-V(x,y)/A},$

Wherein, J ^c(x, y) is the pixel value of the Color Channel correspondence image of image behind the mist elimination, I ^c(x, y) is the original pixel value that the Color Channel correspondence image of mist image I is arranged, and A is the atmosphere light value, V(x, y) be the atmospheric scattering light value;

Step 4.2, with the image of three Color Channels of RGB of the mist elimination image that obtains in the step 4.1, i.e. image J ^R, image J ^GWith image J ^BMake up the image J after namely obtaining restoring.

In the method for the present invention, have mist image I (x, y) to restore to the original of input in the step 1, the algorithm of Main Basis is as follows:

$I(x,y)=J(x,y)(1-\frac{V(x,y)}{A})+V(x,y)$

Wherein, I(x, y) expression is original that mist image, J(x, y arranged) image after expression is restored, V(x, y) expression atmospheric scattering light, A represents whole atmosphere light, this shows needs to use atmosphere light value A and atmospheric scattering light V in the process of restored image.

Embodiment:

Invention provides a kind of simple and effective based on the image/video defogging method capable of dark primary in conjunction with atmospheric scattering light, and effect of the present invention can further specify by following experimental data:

Input the original mist image I that has, original the pixel of mist image is arranged is I(x, y), original the three-component value of mist image RGB is arranged is R=94, G=92, B=96, there is the mist image I to separate with original, extracts and obtain the original image that three Color Channels of RGB of mist image I are arranged, be respectively: image I ^R, image I ^G, image I ^BTo image I ^R, image I ^G, image I ^BCarry out respectively mini-value filtering, obtain image DR, image DG and image DB; More filtered three image DR, DG and DB choose the minimum value of three image corresponding pixel points as the pixel value I of dark primary image corresponding point ^Dark(x, y)=86 namely obtain the original dark primary image I that the mist image I is arranged ^Dark

According to the original dark primary image I that the mist image I is arranged of obtaining ^DarkSeek out the histogram of dark primary image; According to the histogram of dark primary image, in brightness value is front 0.1% dark primary image, seek the pixel value of brightness maximum as threshold value, behind the definite threshold, the selected pixels value is greater than the pixel of threshold value in the histogram of dark primary image; With the pixel chosen corresponding to the original RGB three-component image that the mist image I is arranged, extract respectively the maximal value in the pixel value in the corresponding original RGB three-component image that the mist image I arranged of pixel, namely obtain atmosphere light value A, be respectively airR=237, airG=224 and airB=202;

With the original image that three Color Channels of RGB of mist image I are arranged, i.e. image I ^R, image I ^G, image I ^BIn the minimum value of each pixel respective pixel value extract; Pixel value W(x, y to the pixel that extracts)=92 carry out medium filtering, obtain C(x, y)=median ₄₁(W(x, y)), C(x, y)=107; According to the result who obtains in step 3.1 and the step 3.2, in conjunction with atmospheric scattering light algorithm, namely obtain original atmospheric scattering light value V(x, the y that the mist image I is arranged)=82;

With the original image that three Color Channels of RGB of mist image I are arranged: image I ^R, image I ^GAnd image I ^BRespectively in conjunction with separately atmosphere light value A and atmospheric scattering light value V(x, the y of Color Channel), use Image Restoration Algorithm, ask for obtaining image J(x, y behind the mist elimination) the image J of three Color Channels of RGB ^R(x, y)=18, J ^G(x, y)=15, J ^B(x, y)=23; With image J ^R, image J ^GWith image J ^BOrder stored interleaved according to BGR makes up, the image J after namely obtaining restoring.

Through method of the present invention process original the mist image is arranged and restore after image compare:

Original have most of zone of mist image I that mist is all arranged, do not see the profile details of Chu's object, seek out first the original dark primary image that the mist image I is arranged, because this moment, the integral image color was partially dark, so that the intensity of image is lower, the minimum pixel value that obtains in the image levels off to 0; In addition, owing in the dark primary image existence of mist being arranged, cause atmosphere light to carry out scattering, formed the fuzzy of the bias distortion of color or scenery, be specially required W(x, y) mean value of image and the difference between the standard deviation; The mist elimination restored image that obtains after adopting the inventive method mist elimination to restore, the quality of its image is more original the mist image, and its visibility has obtained raising clearly, and the image that obtains has good effect of visualization.

Claims (7)

1. based on the image/video defogging method capable of dark primary in conjunction with atmospheric scattering light, it is characterized in that, specifically implement according to following steps:

Step 1, with in the original image processing system that mist image I input computing machine arranged, ask for the original dark primary image I that the mist image I is arranged ^Dark

The dark primary image I that step 2, basis are obtained ^Dark, ask for the original atmosphere light value A that three Color Channels of RGB of mist image I are arranged, be respectively airR, airG and airB;

Step 3, ask for original atmospheric scattering light value V(x, the y that the mist image I is arranged);

Step 4, ask for restored image J behind the mist elimination according to original three channel image of RGB, atmosphere light value A and the atmospheric scattering light value V that the mist image I arranged, finally obtain the restored image J behind the mist elimination.

2. according to claim 1ly it is characterized in that based on the image/video defogging method capable of dark primary in conjunction with atmospheric scattering light, described step 1 is specifically implemented according to following steps:

Step 1.1, with in the original image processing system that mist image I input computing machine arranged, it is original that the pixel of mist image is arranged is I(x, y), there is the mist image I to separate with original, extraction obtains the original image that three Color Channels of RGB of mist image I are arranged, and is respectively: image I ^R, image I ^G, image I ^B

Step 1.2, the original image I that three Color Channels of mist image I are arranged to obtaining through step 1.1 ^R, image I ^G, image I ^BCarry out respectively mini-value filtering, namely obtain the image of filtered three Color Channels, be respectively image DR, image DG and image DB;

Filtered three images in step 1.3, the comparison step 1.2: i.e. image DR, image DG and image DB, choose the minimum value of three image corresponding pixel points as the pixel value I of dark primary image corresponding point ^Dark(x, y) namely obtains the original dark primary image I that the mist image I is arranged ^Dark, the original dark primary image I that the mist image I is arranged ^DarkThe value of each pixel is implemented by following algorithm:

$I^{\mathrm{dark}}(x,y)=\underset{c\∈\{R,G,B\}}{\min }\left(\underset{(x,y)\∈\mathrm{\Ω}(x^{\′},y^{\′})}{\min }\left(I^c(x^{\′},y^{\′})\right)\right),$

Wherein, I ^cFor the original Color Channel that the mist image I is arranged, be I ^R, I ^GAnd I ^B, Ω (x ', y ') be the zone of mini-value filtering.

3. according to claim 2 based on the image/video defogging method capable of dark primary in conjunction with atmospheric scattering light, it is characterized in that, described original the mist image I is arranged is coloured image, directly uses the original RGB three-component that the mist image is arranged, and need not to carry out the conversion of color space.

4. according to claim 2ly it is characterized in that based on the image/video defogging method capable of dark primary in conjunction with atmospheric scattering light, described original the mist image I is arranged is the picture that existed in the computing machine or the real time video data of video file or camera collection.

5. according to claim 1ly it is characterized in that based on the image/video defogging method capable of dark primary in conjunction with atmospheric scattering light, described step 2 is specifically implemented according to following steps:

Step 2.1, the original dark primary image I that the mist image I is arranged of obtaining according to step 1.3 ^DarkSeek out the histogram of dark primary image;

Step 2.2, according to the histogram of the dark primary image of step 2.1, in brightness value is front 0.1% dark primary image, seek the pixel value of brightness maximum as threshold value, behind the definite threshold in the histogram of dark primary image the selected pixels value greater than the pixel of threshold value;

Step 2.3, with the pixel chosen in the step 2.2 corresponding to the original RGB three-component image that the mist image I is arranged that obtains in the step 1.1, extract respectively the maximal value in the pixel value in the corresponding original RGB three-component image that the mist image I arranged of pixel, namely obtain atmosphere light value A, be respectively airR, airG and airB.

6. according to claim 1 based on the image/video defogging method capable of dark primary in conjunction with atmospheric scattering light, it is characterized in that, described step 3 is specifically implemented according to following steps:

Step 3.1, with the original image that three Color Channels of RGB of mist image I are arranged that obtains in the step 1.1, i.e. image I ^R, image I ^G, image I ^BIn the minimum value of each pixel respective pixel value extract:

That is: W(x, y)=min{I(x, y);

Step 3.2, to pixel value W(x, the y of the pixel that extracts in the step 3.1) carry out medium filtering, specifically implement according to following algorithm:

C（x，y）=median _sv（W（x，y））

Wherein, C(x, y) be the result of medium filtering, W(x, y) result that obtains for step 3.1, sv is the square window size of using in the median filter;

Step 3.3, according to the result who obtains in step 3.1 and the step 3.2, in conjunction with atmospheric scattering light algorithm, namely obtain original atmospheric scattering light value V(x, the y that the mist image I is arranged), atmospheric scattering light is specifically implemented according to following algorithm:

B（x，y）=C-median _sv（︱W-C︱）（x，y）

V（x，y）=max（min（ρB（x，y），W（x，y）），0）

Wherein, B(x, y) expression W(x, y) local mean value and the difference of local standard deviation, ρ is for taking advantage of sex factor, the intensity of expression recovery, span are between 0.75 to 0.95, sv is that the square window of using in the median filter is big or small, value is 41.

7. according to claim 1 based on the image/video defogging method capable of dark primary in conjunction with atmospheric scattering light, it is characterized in that, described step 4 is specifically implemented according to following steps:

Step 4.1, with the original image that three Color Channels of RGB of mist image I are arranged in the step 1.1: image I ^R, image I ^GAnd image I ^BThe atmosphere light value A of the separately Color Channel that obtains in the integrating step 2 respectively, atmospheric scattering light value V(x, the y that step 3 obtains), use Image Restoration Algorithm, ask for obtaining image J(x, y behind the mist elimination) the image J of three Color Channels of RGB ^c(x, y), the algorithm of the restored image behind the mist elimination is as follows:

$J^c(x,y)=\frac{I^c(x,y)-V(x,y)}{1-V(x,y)/A},$

Wherein, J ^c(x, y) is the pixel value of the Color Channel correspondence image of image behind the mist elimination, I ^c(x, y) is the original pixel value that the Color Channel correspondence image of mist image I is arranged, and A is the atmosphere light value, V(x, y) be the atmospheric scattering light value;

Step 4.2, with the image of three Color Channels of RGB of the mist elimination image that obtains in the step 4.1, i.e. image J ^R, image J ^GWith image J ^BMake up the image J after namely obtaining restoring.

Images