CN107977942A - A kind of restored method of the single image based on multi-focus image fusion - Google Patents
A kind of restored method of the single image based on multi-focus image fusion Download PDFInfo
- Publication number
- CN107977942A CN107977942A CN201711294667.7A CN201711294667A CN107977942A CN 107977942 A CN107977942 A CN 107977942A CN 201711294667 A CN201711294667 A CN 201711294667A CN 107977942 A CN107977942 A CN 107977942A
- Authority
- CN
- China
- Prior art keywords
- mrow
- image
- msub
- groups
- value
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 20
- 230000004927 fusion Effects 0.000 title claims abstract description 16
- 238000001914 filtration Methods 0.000 claims abstract description 24
- 239000003595 mist Substances 0.000 claims abstract description 17
- 230000003044 adaptive effect Effects 0.000 claims abstract description 11
- 230000011218 segmentation Effects 0.000 claims abstract description 5
- 238000002834 transmittance Methods 0.000 claims description 25
- 238000005457 optimization Methods 0.000 claims description 19
- 238000007500 overflow downdraw method Methods 0.000 claims description 16
- 230000008030 elimination Effects 0.000 claims description 15
- 238000003379 elimination reaction Methods 0.000 claims description 15
- 238000012545 processing Methods 0.000 claims description 7
- 238000004458 analytical method Methods 0.000 claims description 6
- 238000010586 diagram Methods 0.000 claims description 6
- 238000013507 mapping Methods 0.000 claims description 6
- 238000005070 sampling Methods 0.000 claims description 6
- 230000001133 acceleration Effects 0.000 claims description 3
- 238000009499 grossing Methods 0.000 claims description 3
- 230000010354 integration Effects 0.000 claims description 3
- 238000013178 mathematical model Methods 0.000 claims description 3
- 238000012544 monitoring process Methods 0.000 abstract description 6
- 238000011161 development Methods 0.000 abstract description 2
- 230000018109 developmental process Effects 0.000 abstract description 2
- 230000000007 visual effect Effects 0.000 abstract description 2
- 238000011084 recovery Methods 0.000 description 4
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000007429 general method 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/40—Image enhancement or restoration using histogram techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/50—Image enhancement or restoration using two or more images, e.g. averaging or subtraction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- 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
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Processing (AREA)
Abstract
A kind of restored method of the single image based on multi-focus image fusion of the present invention, splits the sky areas in single image using histogram statistical features, obtains three groups of overall situation the atmospheric background light;Traditional Laplce's Filtering Model is optimized, a kind of local Laplce's filtering method of border limitation is provided, improves it and handle the alias of dark channel image;Pass through the method for multi-focus image fusion, merge the three groups of fog free images obtained, obtain final Misty Image, the present invention carries out image restoration using the method for fusion, by the histogram information for studying mist image, it is found that the rule of segmentation, sky areas is split using statistical knowledge, transmissivity is optimized, the local Laplce filtering of independent development adaptive boundary limitation of the present invention, the hue and luminance distortion of Misty Image can be made to be greatly improved, the visual effect of human eye improves notable, for road monitoring, the image in greasy weather can effectively be restored, human observer is set to obtain more effective information.
Description
Technical field
The invention belongs to image processing field, more particularly to a kind of recovery side of the single image based on multi-focus image fusion
Method.
Background technology
Traffic monitoring is mainly made of units such as collecting device, transmission equipment and processing equipments at present.Collecting device
Mainly camera, it is therefore an objective to which the information of road and pedestrian are become into image information by sensor.Current monitor equipment is usually suitable
For under preferable unobstructed natural environment, when running into atrocious weather, during such as the greasy weather, the image of camera collection will be hidden
Gear, road monitoring information will have loss.Accordingly, it is capable to be preferably monitored to the road informations of various natural environments, there are ten
Divide important meaning.
The Misty Image restored method of dark primary theory is currently based on, the general method for using direct estimation the atmospheric background light
The value of global the atmospheric background light is obtained, simple smooth operation is carried out to transmittance figure picture, afterwards by dark model, is carried out
The recovery of Misty Image.Although the method for the type can achieve the purpose that recovery, the coarse acquisition to the atmospheric background light, directly
Connecing influences the brightness of image of image restoration;To the depth of the smooth operation degree of transmissivity, regarding for image after recovering is directly affected
Feel effect.Majority of case, halation occurs in region to the image after processing on high, and the brightness of image also has larger loss.
The content of the invention
It is an object of the invention to provide a kind of restored method of the single image based on multi-focus image fusion, schemes for the greasy weather
As being restored, halation phenomenon is avoided, stablizes the recovering quality of image, allows supervisor to obtain more useful informations, can be right
Road conditions carry out more efficient monitoring.
A kind of restored method of the single image based on multi-focus image fusion of the present invention, first, passes through the single width figure to input
The carry out histogram analysis of picture, obtain the threshold value for splitting sky areas in image, the day dead zone according to this threshold value to single image
Domain is split, and obtains the value of three groups of representative overall situation the atmospheric background light;Secondly, three groups are obtained on the basis of this three class value to help secretly
Road image, handles three groups of dark channel images of acquisition with the local Laplce filtering of adaptive boundary limitation, obtains
Transmittance figure picture to after three groups of optimizations;Again, dark primary theoretical model is used for the transmittance figure picture after three groups of optimizations, obtained
Three groups of different mist elimination images are obtained, image co-registration is carried out using multi-focus image fusion method, obtains a complete mist elimination image.
Specifically include the following steps:
Step 1, by carrying out histogram analysis to the single image of input, obtain the threshold for splitting sky areas in image
Value, splits the sky areas of single image according to this threshold value, obtains the value of three groups of representative overall situation the atmospheric background light
The histogram of RGB triple channels is drawn according to the single image of input, at the low order end of histogram seek to the left
First wave crest is looked for, since the wave crest, progressively to the left, finds first trough, which is exactly the segmentation to be found
The threshold value of sky areas;
A1=max (ac),A2=mean (ac),A3=min (ac)
In formula, g (x) is a Gaussian function, hc(x) be input picture a channel image, * represent process of convolution, fc
(x) image after gaussian filtering, f' are representedc(x) be image after gaussian filtering first derivative,It is to scheme after gaussian filtering
The second dervative of picture, acSplit the threshold value of sky areas in each passage of representative image, wherein c represents appointing in RGB triple channels
One passage, A1It is the maximum in three threshold values, A2It is the average in three threshold values, A3It is the minimum value in three threshold values, A1,
A2,A3It is the value of three groups of representative overall situation the atmospheric background light;
Step 2, obtain three groups of dark channel images on the basis of three class value, the part limited with adaptive boundary
Laplce, which filters, handles three groups of dark channel images of acquisition, obtains the transmittance figure picture after three groups of optimization
Step 21, carry out adaptive boundary limitation to input picture, obtains three groups of dark channel images, its formula is as follows:
In formula, ti(x) it is the transmittance figure picture under different global the atmospheric background light,Represent each passage most
Small pixel value, i.e., start the cycle over from first pixel, the pixel value of each passage be compared, obtained minimum value, i.e.,The max pixel value of each passage is represented, i.e.,Ic
(x) image of any passage in tri- passages of RGB, A are representediTo represent three class values of global the atmospheric background light, the value range of i
[1,2,3];
Step 22, carry out guarantor's side smoothing processing, the transmittance figure after acquisition three width optimization with local Laplce filtering
Picture:
In formula,It is the transmittance figure picture after optimizing under different global the atmospheric background light, FLLF is a kind of improved
Local Laplce filtering based on border limitation, in terms of improving the acceleration that place is laplacian pyramid:
Wherein,It is the pyramid of an output,It is the gaussian pyramid of jth layer
Image, a are a mixed coefficints, rj(I) and rj+1(I) be one on pyramid jth layer and the mapping function of+1 layer of jth, lead to
The down-sampling of laplacian pyramid and the principle of up-sampling are crossed, by image down and is amplified to former scale, to figure
It is smooth as guarantor side can be effectively realized, the gaussian pyramid image of front and rear layer can be merged well with this mapping function,
Accelerate the calculating speed of laplacian pyramid;
Step 3, the optimization at three groups transmittance figure picture on the basis of, with dark primary mathematical model handle, obtain three groups
Image after different defoggings, carries out image co-registration using the multi-focus image fusion method of gradient field, will scheme after three groups of different defoggings
As being fused into a complete mist elimination image
Transmittance figure after step 31, the value of the representative overall situation the atmospheric background light obtained according to step 1 and step 2 optimization
Picture, obtains the image after three width defoggings:
In formula, I (x) be input artwork, Ji(x) be obtain three width defoggings after image, i value range [1,2,
3], dt is an adjustment parameter, and value model is [01],It isLower border value;
Step 32, the multi-focus image fusion method progress image co-registration using gradient field, obtain a complete mist elimination image,
Its formula is as follows:
In formula,It is the mist elimination image finally merged, MF represents the multi-focus image fusion method of gradient field.
The multi-focus image fusion method of the gradient field specifically, the image after above-mentioned three width defogging is transformed into YCbCr space,
Each colour space is merged respectively:
The fusion rule in Y spaces:The gradient map (Yg1, Yg2, Yg3) of brightness space is obtained, takes maximum conduct therein
Final brightness space
Nonlinear extension is carried out to it:
Ygn=(Yg-Ygmax)/(Ygmax-Ygmin)×b+c
In formula, Ygmax,YgminIt is the maximum and minimum value in Yg, Ygn is the final luminance picture obtained, wherein,
B, c are constant;
Cb, Cr Space integration rule:
In formula, Cbi,And Cri,Respectively be input three chromatic diagram pictures and its corresponding weight, Cbn and Crn be
Final chromatic diagram picture.
The present invention splits the sky areas in single image using histogram statistical features, obtains three groups global big
Gas bias light;Traditional Laplce's Filtering Model is optimized, provides a kind of local Laplce filtering of border limitation
Method, improves it and handles the alias of dark channel image;By the method for multi-focus image fusion, three groups of fogless figures of acquisition are merged
Picture, obtains final Misty Image.The present invention carries out the recovery of image using the method for fusion, wherein splitting sky areas, adopts
With statistical knowledge, by the histogram information for studying mist image, it was found that the rule of segmentation, the optimization for transmissivity,
The local Laplce filtering of independent development adaptive boundary limitation of the present invention, Misty Image can be made, which to obtain hue and luminance distortion, to be had
Larger raising, the visual effect of human eye improve significantly, for road monitoring, can effectively restore the image in greasy weather, make monitoring
Person obtains more effective information.
Brief description of the drawings
Fig. 1 is the flow chart of the present invention.
The present invention is further described below in conjunction with the drawings and specific embodiments.
Embodiment
As shown in Figure 1, a kind of restored method of the single image based on multi-focus image fusion of the present invention, first, by defeated
The carry out histogram analysis of the single image entered, obtain the threshold value for splitting sky areas in image, according to this threshold value to single width figure
The sky areas of picture is split, and obtains the value of three groups of representative overall situation the atmospheric background light;Secondly, obtained on the basis of this three class value
Take three groups of dark channel images, with adaptive boundary limitation local Laplce filtering to three groups of dark channel images of acquisition into
Row processing, obtains the transmittance figure picture after three groups of optimization;Again, managed for the transmittance figure picture after three groups of optimizations with dark primary
By model, three groups of mist elimination images are obtained, and a complete mist elimination image is fused into by building multi-focus image fusion method,
Specifically include the following steps:
Step 1, by carrying out histogram analysis to the single image of input, obtain the threshold for splitting sky areas in image
Value, splits the sky areas of single image according to this threshold value, obtains the value of three groups of representative overall situation the atmospheric background light
The histogram of RGB triple channels is drawn according to the single image of input, from the low order end (255 value left and right) of histogram
Start to find first wave crest to the left, since the wave crest, progressively to the left, find first trough, which is exactly to want
The threshold value of the segmentation sky areas of searching;
A1=max (ac),A2=mean (ac),A3=min (ac)
In formula, g (x) is a Gaussian function, hc(x) be input picture a channel image, * represent process of convolution, fc
(x) image after gaussian filtering, f' are representedc(x) be image after gaussian filtering first derivative,It is to scheme after gaussian filtering
The second dervative of picture, acSplit the threshold value of sky areas in each passage of representative image, wherein c represents appointing in RGB triple channels
One passage, A1It is the maximum in three threshold values, A2It is the average in three threshold values, A3It is the minimum value in three threshold values, A1,
A2,A3It is the value of three groups of representative overall situation the atmospheric background light;
Step 2, obtain three groups of dark channel images on the basis of three class value, the part limited with adaptive boundary
Laplce, which filters, handles three groups of dark channel images of acquisition, obtains the transmittance figure picture after three groups of optimization
Step 21, carry out adaptive boundary limitation to input picture, obtains three groups of dark channel images, its formula is as follows:
In formula, ti(x) it is the transmittance figure picture under different global the atmospheric background light,Represent each passage most
Small pixel value, i.e., start the cycle over from first pixel, the pixel value of each passage be compared, obtained minimum value, i.e.,The max pixel value of each passage is represented, i.e.,
Ic(x) image of any passage in tri- passages of RGB, A are representediTo represent three class values of global the atmospheric background light, the value model of i
Enclose [1,2,3];
Step 22, carry out guarantor's side smoothing processing, the transmittance figure after acquisition three width optimization with local Laplce filtering
Picture:
In formula,It is the transmittance figure picture after optimizing under different global the atmospheric background light, FLLF is a kind of improved
Local Laplce filtering based on border limitation, in terms of improving the acceleration that place is laplacian pyramid:
Wherein,It is the pyramid of an output,It is the gaussian pyramid of jth layer
Image, a are a mixed coefficints, rj(I) and rj+1(I) be one on pyramid jth layer and the mapping function of+1 layer of jth, lead to
The down-sampling of laplacian pyramid and the principle of up-sampling are crossed, by image down and is amplified to former scale, to figure
It is smooth as guarantor side can be effectively realized, the gaussian pyramid image of front and rear layer can be merged well with this mapping function,
Accelerate the calculating speed of laplacian pyramid;
Step 3, the optimization at three groups transmittance figure picture on the basis of, with dark primary mathematical model handle, obtain three groups
Image after different defoggings, carries out image co-registration using the multi-focus image fusion method of gradient field, will scheme after three groups of different defoggings
As being fused into a complete mist elimination image
Transmittance figure after step 31, the value of the representative overall situation the atmospheric background light obtained according to step 1 and step 2 optimization
Picture, obtains the image after three width defoggings:
In formula, I (x) be input artwork, Ji(x) be obtain three width defoggings after image, i value range [1,2,
3], dt is an adjustment parameter, and value model is [01],It isLower border value;
Step 32, the multi-focus image fusion method progress image co-registration using gradient field, obtain a complete mist elimination image,
Its formula is as follows:
In formula,It is the mist elimination image finally merged, MF represents the multi-focus image fusion method of gradient field.
The multi-focus image fusion method of the gradient field specifically, the image after above-mentioned three width defogging is transformed into YCbCr space,
Each colour space is merged respectively:
The fusion rule in Y spaces:The gradient map (Yg1, Yg2, Yg3) of brightness space is obtained, takes maximum conduct therein
Final brightness space
Nonlinear extension is carried out to it:
Ygn=(Yg-Ygmax)/(Ygmax-Ygmin)×b+c
In formula, Ygmax,YgminIt is the maximum and minimum value in Yg, Ygn is the final luminance picture obtained, wherein, b
=216, c=19;
Cb, Cr Space integration rule:
In formula, Cbi,And Cri,Respectively be input three chromatic diagram pictures and its corresponding weight, Cbn and Crn be
Final chromatic diagram picture.
The above, is only present pre-ferred embodiments, is not intended to limit the scope of the present invention, therefore
Any subtle modifications, equivalent variations and modifications that every technical spirit according to the present invention makees above example, still belong to
In the range of technical solution of the present invention.
Claims (3)
- A kind of 1. restored method of the single image based on multi-focus image fusion, it is characterised in that:First, the single width to input is passed through The carry out histogram analysis of image, obtain the threshold value for splitting sky areas in image, the sky according to this threshold value to single image Region is split, and obtains the value of three groups of representative overall situation the atmospheric background light;Secondly, obtained on the basis of this three class value three groups it is dark Channel image, is handled three groups of dark channel images of acquisition with the local Laplce filtering of adaptive boundary limitation, Obtain the transmittance figure picture after three groups of optimization;Again, dark primary theoretical model is used for the transmittance figure picture after three groups of optimizations, Three groups of different mist elimination images are obtained, image co-registration is carried out using multi-focus image fusion method, obtains a complete mist elimination image.
- A kind of 2. restored method of single image based on multi-focus image fusion according to claim 1, it is characterised in that tool Body includes the following steps:Step 1, by carrying out histogram analysis to the single image of input, obtain the threshold value for splitting sky areas in image, root Threshold value splits the sky areas of single image accordingly, obtains the value of three groups of representative overall situation the atmospheric background lightThe histogram of RGB triple channels is drawn according to the single image of input, at the low order end of histogram find the to the left One wave crest, since the wave crest, progressively to the left, finds first trough, which is exactly the segmentation sky to be found The threshold value in region;<mfenced open = "{" close = ""> <mtable> <mtr> <mtd> <mrow> <msub> <mi>f</mi> <mi>c</mi> </msub> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> <mo>=</mo> <msub> <mi>h</mi> <mi>c</mi> </msub> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> <mo>*</mo> <mi>g</mi> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> </mrow> </mtd> </mtr> <mtr> <mtd> <mrow> <msub> <mi>a</mi> <mi>c</mi> </msub> <mo>=</mo> <munder> <mi>argmax</mi> <mrow> <mi>x</mi> <mo>&Element;</mo> <mo>&lsqb;</mo> <mn>0</mn> <mo>,</mo> <mn>255</mn> <mo>&rsqb;</mo> </mrow> </munder> <mrow> <mo>(</mo> <mi>x</mi> <mo>|</mo> <msubsup> <mi>f</mi> <mi>c</mi> <mo>&prime;</mo> </msubsup> <mo>(</mo> <mi>x</mi> <mo>)</mo> <mo>=</mo> <mn>0</mn> <mo>,</mo> <msubsup> <mi>f</mi> <mi>c</mi> <mrow> <mo>&prime;</mo> <mo>&prime;</mo> </mrow> </msubsup> <mo>(</mo> <mi>x</mi> <mo>)</mo> <mo>></mo> <mn>0</mn> <mo>)</mo> </mrow> <mo>,</mo> <mi>c</mi> <mo>&Element;</mo> <mo>{</mo> <mi>r</mi> <mo>,</mo> <mi>g</mi> <mo>,</mo> <mi>b</mi> <mo>}</mo> </mrow> </mtd> </mtr> </mtable> </mfenced>A1=max (ac), A2=mean (ac), A3=min (ac)In formula, g (x) is a Gaussian function, hc(x) be input picture a channel image, * represent process of convolution, fc(x) Represent the image after gaussian filtering, f 'c(x) be image after gaussian filtering first derivative, f "c(x) it is image after gaussian filtering Second dervative, acSplit the threshold value of sky areas in each passage of representative image, wherein c represents any logical in RGB triple channels Road, A1It is the maximum in three threshold values, A2It is the average in three threshold values, A3It is the minimum value in three threshold values, A1, A2, A3 It is the value of three groups of representative overall situation the atmospheric background light;Step 2, obtain three groups of dark channel images on the basis of three class value, is drawn with the part of adaptive boundary limitation general Lars, which filters, handles three groups of dark channel images of acquisition, obtains the transmittance figure picture after three groups of optimizationStep 21, carry out adaptive boundary limitation to input picture, obtains three groups of dark channel images, its formula is as follows:<mrow> <msub> <mi>t</mi> <mi>i</mi> </msub> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> <mo>=</mo> <munder> <mi>min</mi> <mrow> <mi>i</mi> <mo>&Element;</mo> <mo>&lsqb;</mo> <mn>1</mn> <mo>,</mo> <mn>2</mn> <mo>,</mo> <mn>3</mn> <mo>&rsqb;</mo> </mrow> </munder> <mo>{</mo> <munder> <mi>max</mi> <mrow> <mi>c</mi> <mo>&Element;</mo> <mo>&lsqb;</mo> <mi>r</mi> <mo>,</mo> <mi>g</mi> <mo>,</mo> <mi>b</mi> <mo>&rsqb;</mo> </mrow> </munder> <mrow> <mo>(</mo> <mfrac> <mrow> <msub> <mi>A</mi> <mi>i</mi> </msub> <mo>-</mo> <msub> <mi>I</mi> <mi>c</mi> </msub> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> </mrow> <mrow> <msub> <mi>A</mi> <mi>i</mi> </msub> <mo>-</mo> <msubsup> <mi>C</mi> <mn>0</mn> <mi>c</mi> </msubsup> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> </mrow> </mfrac> <mo>,</mo> <mfrac> <mrow> <msub> <mi>A</mi> <mi>i</mi> </msub> <mo>-</mo> <msub> <mi>I</mi> <mi>c</mi> </msub> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> </mrow> <mrow> <msub> <mi>A</mi> <mi>i</mi> </msub> <mo>-</mo> <msubsup> <mi>C</mi> <mn>1</mn> <mi>c</mi> </msubsup> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> </mrow> </mfrac> <mo>)</mo> </mrow> <mo>}</mo> </mrow>In formula, ti(x) it is the transmittance figure picture under different global the atmospheric background light,Represent the minimum image of each passage Element value, i.e., start the cycle over from first pixel, the pixel value of each passage be compared, obtained minimum value, i.e., The max pixel value of each passage is represented, i.e., Ic(x) image of any passage in tri- passages of RGB, A are representediTo represent three class values of global the atmospheric background light, the value model of i Enclose [1,2,3];Step 22, carry out guarantor's side smoothing processing, the transmittance figure picture after acquisition three width optimization with local Laplce filtering:In formula,It is the transmittance figure picture after optimizing under different global the atmospheric background light, FLLF, which is that one kind is improved, to be based on The local Laplce filtering of border limitation, in terms of improving the acceleration that place is laplacian pyramid:<mrow> <msub> <mi>L</mi> <msub> <mi>l</mi> <mn>0</mn> </msub> </msub> <mo>&lsqb;</mo> <mn>0</mn> <mo>&rsqb;</mo> <mrow> <mo>(</mo> <mi>x</mi> <mo>,</mo> <mi>y</mi> <mo>)</mo> </mrow> <mo>=</mo> <mrow> <mo>(</mo> <mn>1</mn> <mo>-</mo> <mi>&alpha;</mi> <mo>)</mo> </mrow> <msub> <mi>G</mi> <msub> <mi>l</mi> <mn>0</mn> </msub> </msub> <mo>&lsqb;</mo> <msub> <mi>r</mi> <mi>j</mi> </msub> <mrow> <mo>(</mo> <mi>I</mi> <mo>)</mo> </mrow> <mo>&rsqb;</mo> <mrow> <mo>(</mo> <mi>x</mi> <mo>,</mo> <mi>y</mi> <mo>)</mo> </mrow> <mo>+</mo> <msub> <mi>&alpha;L</mi> <msub> <mi>l</mi> <mn>0</mn> </msub> </msub> <mo>&lsqb;</mo> <msub> <mi>r</mi> <mrow> <mi>j</mi> <mo>+</mo> <mn>1</mn> </mrow> </msub> <mrow> <mo>(</mo> <mi>I</mi> <mo>)</mo> </mrow> <mo>&rsqb;</mo> <mrow> <mo>(</mo> <mi>x</mi> <mo>,</mo> <mi>y</mi> <mo>)</mo> </mrow> </mrow>Wherein,It is the pyramid of an output,It is the gaussian pyramid image of jth layer, A is a mixed coefficint, rj(I) and rj+1(I) be one on pyramid jth layer and the mapping function of+1 layer of jth, pass through drawing The principle of this pyramidal down-sampling of pula and up-sampling, by image down and is amplified to former scale, can to image It is smooth to effectively realize guarantor side, the gaussian pyramid image of front and rear layer can be merged well with this mapping function, accelerated The calculating speed of laplacian pyramid;Step 3, the optimization at three groups transmittance figure picture on the basis of, handled with dark primary mathematical model, obtain three groups it is different Defogging after image, using gradient field multi-focus image fusion method carry out image co-registration, image after three groups of different defoggings is melted Synthesize a complete mist elimination imageTransmittance figure picture after step 31, the value of the representative overall situation the atmospheric background light obtained according to step 1 and step 2 optimization, is obtained Take the image after three width defoggings:<mrow> <msub> <mi>J</mi> <mi>i</mi> </msub> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> <mo>=</mo> <mrow> <mo>(</mo> <mi>I</mi> <mo>(</mo> <mi>x</mi> <mo>)</mo> <mo>-</mo> <msub> <mi>A</mi> <mi>i</mi> </msub> <mo>)</mo> </mrow> <mo>/</mo> <msup> <mrow> <mo>(</mo> <mi>m</mi> <mi>a</mi> <mi>x</mi> <mo>(</mo> <mrow> <msub> <mover> <mi>t</mi> <mo>~</mo> </mover> <mi>i</mi> </msub> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> <mo>,</mo> <msub> <mover> <mi>t</mi> <mo>~</mo> </mover> <mn>0</mn> </msub> </mrow> <mo>)</mo> <mo>)</mo> </mrow> <mrow> <mi>d</mi> <mi>t</mi> </mrow> </msup> <mo>+</mo> <msub> <mi>A</mi> <mi>i</mi> </msub> <mo>,</mo> <mi>i</mi> <mo>&Element;</mo> <mo>{</mo> <mn>1</mn> <mo>,</mo> <mn>2</mn> <mo>,</mo> <mn>3</mn> <mo>}</mo> </mrow>In formula, I (x) be input artwork, Ji(x) be obtain three width defoggings after image, the value range [1,2,3] of i, dt It is an adjustment parameter, value model is [01],It isLower border value;Step 32, the multi-focus image fusion method progress image co-registration using gradient field, obtain a complete mist elimination image, it is public Formula is as follows:<mrow> <mover> <mi>J</mi> <mo>~</mo> </mover> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> <mo>=</mo> <munder> <mrow> <mi>M</mi> <mi>F</mi> </mrow> <mrow> <mi>i</mi> <mo>&Element;</mo> <mo>&lsqb;</mo> <mn>1</mn> <mo>,</mo> <mn>2</mn> <mo>,</mo> <mn>3</mn> <mo>&rsqb;</mo> </mrow> </munder> <mrow> <mo>(</mo> <msub> <mi>J</mi> <mi>i</mi> </msub> <mo>(</mo> <mi>x</mi> <mo>)</mo> <mo>)</mo> </mrow> <mo>,</mo> </mrow>In formula,It is the mist elimination image finally merged, MF represents the multi-focus image fusion method of gradient field.
- 3. the restored method of a kind of single image based on multi-focus image fusion according to claim 2, it is characterised in that should The multi-focus image fusion method of gradient field by the image after above-mentioned three width defogging specifically, be transformed into YCbCr space, respectively to each The colour space is merged:The fusion rule in Y spaces:The gradient map (Yg1, Yg2, Yg3) of brightness space is obtained, takes maximum therein as final Brightness spaceNonlinear extension is carried out to it:Ygn=(Yg-Ygmax)/(Ygmax-Ygmin)×b+cIn formula, Ygmax, YgminIt is the maximum and minimum value in Yg, Ygn is the final luminance picture obtained, wherein, b, c are Constant;Cb, Cr Space integration rule:<mfenced open = "{" close = ""> <mtable> <mtr> <mtd> <mrow> <mi>C</mi> <mi>b</mi> <mi>n</mi> <mo>=</mo> <munderover> <mo>&Sigma;</mo> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mn>3</mn> </munderover> <mrow> <msubsup> <mi>&omega;</mi> <mi>b</mi> <mi>i</mi> </msubsup> <mo>&CenterDot;</mo> <msup> <mi>Cb</mi> <mi>i</mi> </msup> <mo>,</mo> <msubsup> <mi>where&omega;</mi> <mi>b</mi> <mi>i</mi> </msubsup> <mo>=</mo> <msup> <mi>Cb</mi> <mi>i</mi> </msup> <mo>/</mo> <munderover> <mo>&Sigma;</mo> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mn>3</mn> </munderover> <mrow> <msup> <mi>Cb</mi> <mi>i</mi> </msup> </mrow> </mrow> </mrow> </mtd> </mtr> <mtr> <mtd> <mrow> <mi>C</mi> <mi>r</mi> <mi>n</mi> <mo>=</mo> <munderover> <mo>&Sigma;</mo> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mn>3</mn> </munderover> <mrow> <msubsup> <mi>&omega;</mi> <mi>r</mi> <mi>i</mi> </msubsup> <mo>&CenterDot;</mo> <msup> <mi>Cr</mi> <mi>i</mi> </msup> <mo>,</mo> <msubsup> <mi>where&omega;</mi> <mi>r</mi> <mi>i</mi> </msubsup> <mo>=</mo> <msup> <mi>Cr</mi> <mi>i</mi> </msup> <mo>/</mo> <munderover> <mo>&Sigma;</mo> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mn>3</mn> </munderover> <mrow> <msup> <mi>Cr</mi> <mi>i</mi> </msup> </mrow> </mrow> </mrow> </mtd> </mtr> </mtable> </mfenced>In formula, Cbi,And Cri,It is three chromatic diagram pictures of input and its corresponding weight respectively, Cbn and Crn are final Chromatic diagram picture.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711294667.7A CN107977942B (en) | 2017-12-08 | 2017-12-08 | Single image restoration method based on multi-focus fusion |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711294667.7A CN107977942B (en) | 2017-12-08 | 2017-12-08 | Single image restoration method based on multi-focus fusion |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107977942A true CN107977942A (en) | 2018-05-01 |
CN107977942B CN107977942B (en) | 2020-06-05 |
Family
ID=62009513
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201711294667.7A Active CN107977942B (en) | 2017-12-08 | 2017-12-08 | Single image restoration method based on multi-focus fusion |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107977942B (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110211072A (en) * | 2019-06-11 | 2019-09-06 | 青岛大学 | A kind of image defogging method, system and electronic equipment and storage medium |
CN110533616A (en) * | 2019-08-30 | 2019-12-03 | 福建省德腾智能科技有限公司 | A kind of method of image sky areas segmentation |
CN110619633A (en) * | 2019-09-10 | 2019-12-27 | 武汉科技大学 | Liver image segmentation method based on multi-path filtering strategy |
CN110648302A (en) * | 2019-10-08 | 2020-01-03 | 太原科技大学 | Light field full-focus image fusion method based on edge enhancement guide filtering |
CN111866345A (en) * | 2020-07-28 | 2020-10-30 | 杭州武盛广告制作有限公司 | Image acquisition device for video monitoring system suitable for extreme weather |
CN111932469A (en) * | 2020-07-21 | 2020-11-13 | 泉州职业技术大学 | Significance weight quick exposure image fusion method, device, equipment and medium |
CN111932470A (en) * | 2020-07-21 | 2020-11-13 | 泉州职业技术大学 | Image restoration method, device, equipment and medium based on visual selection fusion |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101980290A (en) * | 2010-10-29 | 2011-02-23 | 西安电子科技大学 | Method for fusing multi-focus images in anti-noise environment |
CN105631823A (en) * | 2015-12-28 | 2016-06-01 | 西安电子科技大学 | Dark channel sky area defogging method based on threshold segmentation optimization |
US20160328789A1 (en) * | 2012-09-28 | 2016-11-10 | Interactive Memories, Inc. | Methods and systems for automatically searching for related digital templates during media-based project creation |
CN106875371A (en) * | 2017-02-09 | 2017-06-20 | 聚龙智瞳科技有限公司 | Image interfusion method and image fusion device based on Bayer format |
CN106940882A (en) * | 2017-02-15 | 2017-07-11 | 国网江苏省电力公司常州供电公司 | A kind of transformer substation video image clarification method for meeting human-eye visual characteristic |
-
2017
- 2017-12-08 CN CN201711294667.7A patent/CN107977942B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101980290A (en) * | 2010-10-29 | 2011-02-23 | 西安电子科技大学 | Method for fusing multi-focus images in anti-noise environment |
US20160328789A1 (en) * | 2012-09-28 | 2016-11-10 | Interactive Memories, Inc. | Methods and systems for automatically searching for related digital templates during media-based project creation |
CN105631823A (en) * | 2015-12-28 | 2016-06-01 | 西安电子科技大学 | Dark channel sky area defogging method based on threshold segmentation optimization |
CN106875371A (en) * | 2017-02-09 | 2017-06-20 | 聚龙智瞳科技有限公司 | Image interfusion method and image fusion device based on Bayer format |
CN106940882A (en) * | 2017-02-15 | 2017-07-11 | 国网江苏省电力公司常州供电公司 | A kind of transformer substation video image clarification method for meeting human-eye visual characteristic |
Non-Patent Citations (4)
Title |
---|
董蕊 等: "基于统计直方图的单幅图像去雾算法", 《计算机应用与软件》 * |
高银 等: "基于TV模型的暗原色理论雾天图像复原算法", 《中国激光》 * |
高银: "雾天彩***图像的清晰化算法研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
黄倩文: "基于三通道不可分对称小波的多聚焦图像融合", 《湖北大学学报》 * |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110211072A (en) * | 2019-06-11 | 2019-09-06 | 青岛大学 | A kind of image defogging method, system and electronic equipment and storage medium |
CN110211072B (en) * | 2019-06-11 | 2023-05-02 | 青岛大学 | Image defogging method and system, electronic equipment and storage medium |
CN110533616A (en) * | 2019-08-30 | 2019-12-03 | 福建省德腾智能科技有限公司 | A kind of method of image sky areas segmentation |
CN110619633A (en) * | 2019-09-10 | 2019-12-27 | 武汉科技大学 | Liver image segmentation method based on multi-path filtering strategy |
CN110619633B (en) * | 2019-09-10 | 2023-06-23 | 武汉科技大学 | Liver image segmentation method based on multipath filtering strategy |
CN110648302A (en) * | 2019-10-08 | 2020-01-03 | 太原科技大学 | Light field full-focus image fusion method based on edge enhancement guide filtering |
CN110648302B (en) * | 2019-10-08 | 2022-04-12 | 太原科技大学 | Light field full-focus image fusion method based on edge enhancement guide filtering |
CN111932469A (en) * | 2020-07-21 | 2020-11-13 | 泉州职业技术大学 | Significance weight quick exposure image fusion method, device, equipment and medium |
CN111932470A (en) * | 2020-07-21 | 2020-11-13 | 泉州职业技术大学 | Image restoration method, device, equipment and medium based on visual selection fusion |
CN111932469B (en) * | 2020-07-21 | 2024-05-17 | 泉州职业技术大学 | Method, device, equipment and medium for fusing saliency weight fast exposure images |
CN111866345A (en) * | 2020-07-28 | 2020-10-30 | 杭州武盛广告制作有限公司 | Image acquisition device for video monitoring system suitable for extreme weather |
Also Published As
Publication number | Publication date |
---|---|
CN107977942B (en) | 2020-06-05 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107977942A (en) | A kind of restored method of the single image based on multi-focus image fusion | |
CN105046666B (en) | A kind of method of the real-time defogging of traffic video based on dark primary priori | |
CN102750674B (en) | Video image defogging method based on self-adapting allowance | |
CN104050637B (en) | Quick image defogging method based on two times of guide filtration | |
CN103049888A (en) | Image/video demisting method based on combination of dark primary color of atmospheric scattered light | |
CN103020914B (en) | Based on the rapid image defogging method capable of spatial continuity principle | |
CN104036466B (en) | A kind of video defogging method and system | |
CN102243758A (en) | Fog-degraded image restoration and fusion based image defogging method | |
CN107578386A (en) | A kind of optimization defogging processing method of unmanned plane shooting image | |
CN105976337B (en) | A kind of image defogging method based on intermediate value guiding filtering | |
CN102254313A (en) | Image defogging method based on restoration and fusion of images on foggy days | |
CN102831591A (en) | Gaussian filter-based real-time defogging method for single image | |
CN105354806A (en) | Dark channel based rapid defogging method and system | |
CN104318524A (en) | Method, device and system for image enhancement based on YCbCr color space | |
CN105096278A (en) | Image enhancement method based on illumination adjustment and equipment thereof | |
CN104766286A (en) | Image defogging device and method based on pilotless automobile | |
CN103914820A (en) | Image haze removal method and system based on image layer enhancement | |
CN107146209A (en) | A kind of single image to the fog method based on gradient field | |
CN105913390A (en) | Image defogging method and system | |
CN105931208A (en) | Physical model-based low-illuminance image enhancement algorithm | |
CN105701783A (en) | Single image defogging method based on ambient light model and apparatus thereof | |
CN104318528A (en) | Foggy weather image restoration method based on multi-scale WLS filtering | |
CN103778605A (en) | Greasy weather image enhancement method | |
CN112164010A (en) | Multi-scale fusion convolution neural network image defogging method | |
CN102968767A (en) | Method for real-time restoration of fog-degraded image with white balance correction |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |