CN105787896B - A kind of light and color homogenization method and system between the image for taking radiation two-dimensional distribution into account - Google Patents
A kind of light and color homogenization method and system between the image for taking radiation two-dimensional distribution into account Download PDFInfo
- Publication number
- CN105787896B CN105787896B CN201610111326.0A CN201610111326A CN105787896B CN 105787896 B CN105787896 B CN 105787896B CN 201610111326 A CN201610111326 A CN 201610111326A CN 105787896 B CN105787896 B CN 105787896B
- Authority
- CN
- China
- Prior art keywords
- mrow
- image
- msubsup
- msub
- radiation
- 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.)
- Active
Links
- 230000005855 radiation Effects 0.000 title claims abstract description 57
- 238000000034 method Methods 0.000 title claims abstract description 41
- 238000000265 homogenisation Methods 0.000 title claims abstract description 26
- 230000009466 transformation Effects 0.000 claims abstract description 23
- 238000012545 processing Methods 0.000 claims abstract description 17
- 238000009499 grossing Methods 0.000 claims abstract description 13
- 238000000844 transformation Methods 0.000 claims abstract description 8
- 238000001914 filtration Methods 0.000 claims description 22
- 238000006243 chemical reaction Methods 0.000 claims description 9
- 238000012937 correction Methods 0.000 claims description 5
- 230000000007 visual effect Effects 0.000 abstract description 2
- 238000003672 processing method Methods 0.000 abstract 1
- 238000005286 illumination Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 238000003384 imaging method Methods 0.000 description 3
- 238000005070 sampling Methods 0.000 description 2
- 238000007792 addition Methods 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000009931 harmful effect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- VMXUWOKSQNHOCA-UKTHLTGXSA-N ranitidine Chemical compound [O-][N+](=O)\C=C(/NC)NCCSCC1=CC=C(CN(C)C)O1 VMXUWOKSQNHOCA-UKTHLTGXSA-N 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 238000010129 solution processing Methods 0.000 description 1
- 230000006641 stabilisation Effects 0.000 description 1
- 238000011105 stabilization Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Classifications
-
- G06T5/70—
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformation in the plane of the image
- G06T3/40—Scaling the whole image or part thereof
- G06T3/4038—Scaling the whole image or part thereof for image mosaicing, i.e. plane images composed of plane sub-images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10032—Satellite or aerial image; Remote sensing
-
- 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
- G06T2207/20172—Image enhancement details
- G06T2207/20182—Noise reduction or smoothing in the temporal domain; Spatio-temporal filtering
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
- Color Image Communication Systems (AREA)
Abstract
A kind of light and color homogenization method and system between the image for taking radiation two-dimensional distribution into account, the object level smoothing processing removal excessively bright and excessively dark target of prospect is carried out including treating light and color homogenization image, average statistical and variance carry out radiation adjustment using whole statistical parameter to every image after smooth;Pre-splicing obtain with reference to low frequency image is carried out to radiation adjustment result;The low-and high-frequency part after contourlet transformation is separated is carried out, radiation adjustment is carried out to low frequency part using differential technique, Contourlet inverse transformations is carried out and obtains the image after individual radiation adjustment;Image after all radiation adjustment is spliced to obtain final light and color homogenization result splicing image.Processing method of the present invention is clear, it is workable, brightness tones inconsistence problems between different times image can be not only eliminated, while can also solve Luminance Distribution inconsistence problems caused by image is limited in itself due to sensor so that final image joint result has better visual effect.
Description
Technical field
The present invention relates to Remote Sensing Image Processing Technology field, more particularly, between a kind of image for taking radiation two-dimensional distribution into account
Light and color homogenization method and system.
Background technology
Remote sensing image causes shadow due to the difference of the image-forming conditions such as the time of acquisition, illumination, camera angle and type of ground objects
There are the uneven illuminations phenomenon such as Luminance Distribution is uneven, contrast is uneven and color is inconsistent as between, so as to greatly restrict
The using effect of remote sensing image in follow-up orthography splicing and other image engineerings applications.Even color is exactly to eliminate between image
Therefore these non-uniform processes, in order to which more accurate and real surface reaches the objective reality world, obtain the image information of high quality
And be effectively used, light and color homogenization is very necessary.
Unmanned plane imaging technique is quickly grown in recent years, fast and flexible, can obtain that quantity is larger but breadth the short time
The characteristics of small image, so that application scenarios are more and more wider.Since unmanned plane image height is low, platform stabilization is poor, therefore takes the photograph
Shadow angle change is larger, and same atural object can be compared with due to the hue and luminance differences that are reflected on image of difference of camera angle
Greatly;If sun altitude is relatively low during imaging, since the influence of shade causes image sunny slope apparent luminance difference occur with opaco
Different, showing on the whole between different course lines has apparent dark bands phenomenon;And the quality limitation of ordinary digital camera causes
The Luminance Distribution of individual image in itself is uneven, and it is partially dark to typically appear as 4 jiaos of image;It is had for the larger survey area of coverage
Different time imaging is possible, and the difference of different time weather can so that more phase image global illumination differences are apparent, to subsequently splicing
As a result application has some impact on.
The content of the invention
In view of the above-mentioned problems, the present invention proposes the light and color homogenization technical side between a kind of image for taking radiation two-dimensional distribution into account
Case, it is workable, brightness tones inconsistence problems between different times image can be not only eliminated, while can also solve image sheet
Luminance Distribution inconsistence problems caused by body is limited due to sensor so that final image joint result has better vision to imitate
Fruit.
The technical scheme is that a kind of light and color homogenization method between image for taking radiation two-dimensional distribution into account, including following
Step:
Step 1, if shared N image I to be splicedi, i=1,2,3...N, remove individual using the smoothing method of object level
Image I to be splicediIn highlighted or excessively dark foreground target, obtain object level it is smooth after image SIi, i=1,2,3...N;
Step 2, count all object levels it is smooth after image SIiAverage SIimWith variance SIiv, obtain smooth rear whole
Average m and variance v;
Step 3, according to average m and variance v, it is smooth to each object level after image SIiCarry out the spoke based on mean variance
Adjustment is penetrated, obtains radiation adjustment result image Ii', i=1,2,3...N;
Step 4, to all N radiation adjustment result image Ii' progress is pre-splicing, and pre-splicing result is carried out smoothly to obtain
Low-frequency information image B is referred to whole;
Step 5, to Ii、SIiAnd BIiContourlet transformation is carried out respectively obtains corresponding low frequency partAnd high frequency sectionWherein BIiFor integrally with reference to basis on low-frequency information image B
Image I to be splicediThe image that correspondence position is cut, i=1,2,3...N;
Step 6, it is rightWithFurther low-pass filtering treatment is carried out, obtains low-pass filtering resultWithIt utilizes
The low-pass filtering result of the conversion low frequency part of acquisitionAndIt is adjusted using difference or ratio formula
Result afterwardsI=1,2,3...N;
The difference value equation is as follows,
The ratio formula is as follows,
Step 7, utilizeIt replacesAnd combine original high frequency sectionContourlet inverse transformations are carried out, are obtained everywhere
Image ZI after even color processing after reasoni, i=1,2,3...N;
Step 8, to the image ZI after all N radiation adjustmentiSpliced to obtain final light and color homogenization result splicing shadow
As Z.
Moreover, in step 3, the radiation adjustment realization method based on mean variance is as follows,
IfFor SIiKth wave band average gray,It is the average for the whole kth wave band that statistics obtains,WithIt is the standard deviation of corresponding wave band, mean variance method radiant correction coefficient is:
Then I 'iKth wave band be I 'ik=akSIik+bk。
The present invention correspondingly provides the light and color homogenization system between a kind of image for taking radiation two-dimensional distribution into account, including with lower die
Block:
First module, for setting shared N image I to be splicedi, i=1,2,3...N, utilize the smoothing method of object level
Remove individual image I to be splicediIn highlighted or excessively dark foreground target, obtain object level it is smooth after image SIi, i=1,2,
3...N;
Second module, for count all object levels it is smooth after image SIiAverage SIimWith variance SIiv, obtain smooth
Whole average m and variance v afterwards;
3rd module, for according to average m and variance v, it is smooth to each object level after image SIiIt carries out being based on average side
The radiation adjustment of difference obtains radiation adjustment result image Ii', i=1,2,3...N;
4th module, for all N radiation adjustment result image Ii' carry out it is pre-splicing, and to pre-splicing result into
Row smoothly obtains whole with reference to low-frequency information image B;
5th module, for Ii、SIiAnd BIiContourlet transformation is carried out respectively obtains corresponding low frequency part And high frequency sectionWherein BIiIntegrally to be treated with reference to basis on low-frequency information image B
Splice image IiThe image that correspondence position is cut, i=1,2,3...N;
6th module, for pairWithFurther low-pass filtering treatment is carried out, obtains low-pass filtering resultWithUtilize the low-pass filtering result of the conversion low frequency part of acquisitionAndIt is public using difference or ratio
Formula be adjusted after resultI=1,2,3...N;
The difference value equation is as follows,
The ratio formula is as follows,
7th module, for utilizingIt replacesAnd combine original high frequency sectionCarry out Contourlet inversions
It changes, image ZI after the even color processing that obtains that treatedi, i=1,2,3...N;
8th module, for the image ZI after all N radiation adjustmentiSpliced to obtain final light and color homogenization result
Splice image Z.
Moreover, in the 3rd module, the radiation adjustment realization method based on mean variance is as follows,
IfFor SIiKth wave band average gray,It is the average for the whole kth wave band that statistics obtains,WithIt is the standard deviation of corresponding wave band, mean variance method radiant correction coefficient is:
Then I 'iKth wave band be I 'ik=akSIik+bk。
Technical solution processing method provided by the present invention is clear, workable, can not only eliminate different times image
Between brightness tones inconsistence problems, while can also solve that Luminance Distribution is inconsistent asks caused by image is limited in itself due to sensor
Topic so that final image joint result has better visual effect.
Description of the drawings
Fig. 1 is the flow chart of the embodiment of the present invention.
Fig. 2 is the image joint figure before the light and color homogenization of the embodiment of the present invention.
Fig. 3 is the image joint figure after the light and color homogenization of the embodiment of the present invention.
Specific embodiment
Light and color homogenization method between a kind of image for taking radiation two-dimensional distribution into account provided by the present invention is that treat even light even
Color image carries out object level smoothing processing and obtains entirety with reference to low frequency;Then height frequency division is carried out using contourlet transformation
From;It still can be there are a part of high-frequency information, so needing due to having carried out the low frequency part after multistage contourlet transformation
Low frequency part after background radiation difference is added to the conversion of pending image will be obtained and finally carry out Contourlet inverse transformations;Most
Whole even color processing and splicing are carried out afterwards.When it is implemented, computer software technology, which can be used, realizes automatic flow operation, below
Technical solution that the present invention will be described in detail in conjunction with the accompanying drawings and embodiments.
As shown in Figure 1, the flow of embodiment specifically includes following steps:
Step 1, treat light and color homogenization image and carry out the object level smoothing processing removal excessively bright and excessively dark target of prospect, obtain generation
The smooth rear image of table image real background radiation information:
Assuming that the image I of common N even color splicingsi(i=1,2,3...N), the whole low-frequency information image that refers to is B, is handled
Process may be summarized to be:Individual image I to be spliced is removed using the smoothing method of object leveliIn highlighted or excessively dark prospect mesh
Mark, obtains SIi(i=1,2,3...N);
The purpose of even color is so that image to be spliced is due to different caused by image-forming condition and time difference in image joint
Hue and luminance be finally consistent on the whole, and the Detail contrast in holding image itself and real contrast.Therefore
The essential step acquisition of entire even color processing can represent the background radiation information of whole splicing result.When it is implemented, object
Grade smoothing processing specific implementation can be found in document:Li,Wenzhuo,Kaimin Sun,and Hongya Zhang."
Algorithm for relative radiometric consistency process of remote sensing
images based on object-oriented smoothing and contourlet transforms."Journal
of Applied Remote Sensing 8.1(2014):083607-083607.
Itself there is highlighted target or excessively dark target in individual image, this partial target belongs to the prospect of image, it is necessary to will
Such target is excluded outside the process obtained in whole background radiation information.Therefore obtain and represent image real background radiation letter
Breath.
Step 2, average and variance statistic are carried out to the image after smooth, obtains the whole statistical parameter of multiple images:
Count all object levels it is smooth after image SIiAverage SIiMWith variance SIiVObtain smooth rear whole average m
With variance v;
Average value standard deviation statistics is carried out to the above-mentioned all images for carrying out prospect removal, obtains statistical parameter.
Step 3, using whole statistical parameter, (average m and variance v) carries out based on average side every image after smooth
The radiation adjustment of difference, obtain object level it is smooth after mean variance consistent (average value standard deviation adjustment) radiation adjustment result
Image:
To SIiIt carries out obtaining radiation adjustment result image I based on average value standard deviation radiation consistency treatmenti', i=1,2,
3...N。
Detailed process is as follows:Assuming thatFor SIiKth wave band average gray,It is the whole kth ripple that statistics obtains
The average of section,WithIt is the standard deviation of corresponding wave band.So mean variance method radiant correction coefficient is:
That is I 'iKth wave band be:I′ik=akSIik+bk
Step 4, it is pre-splicing to these radiation adjustment result images progress, and carry out Gaussian smoothing and obtain representing whole background
The consistent reference low frequency image of radiation:
To Ii' progress is pre-splicing, and pre-splicing result is carried out smoothly to obtain whole with reference to low-frequency information image B.
After the average value standard deviation based on statistics is carried out to individual image and is adjusted, these images are carried out pre-splicing
It connects, causes that some apparent splicing traces, these edges letter may be had in the process of splicing due to choosing splicing line
Breath may be mixed into pending image in subsequent processing, it is therefore desirable to smooth, guarantee entirety is carried out to pre-splicing result
With reference to the acquisition reliability of low-frequency information.The corresponding prior art can be used in specific pre-splicing and progress Gaussian smoothing, and the present invention is not
It gives and repeating.
Step 5, to every raw video, the smooth image after foreground target and the reference low frequency of its corresponding region are removed
Image carries out the low-and high-frequency part after contourlet transformation is separated:
To Ii、SIiAnd BIiN times contourlet transformation is carried out respectively obtains corresponding low frequency part
And high frequency sectionWherein BIiIntegrally to refer on low-frequency information image B according to image I to be splicediIt is right
The image that position is cut is answered, i.e. corresponding region is local;When it is implemented, those skilled in the art can sets itself n take
Value suggests that n takes 4 according to result of the test.
To every raw video, remove the smooth image after foreground target and it is corresponding in the splicing result of step 4
When the reference low frequency image in region carries out the low-and high-frequency part after contourlet transformation (contourlet transform) is separated,
Contourlet transformation specific implementation can be found in pertinent literature, for ease of implementing to refer to, provide respective description.Contourlet becomes
The decomposable process changed can be summarised as following steps:
It is adopted 1. a pair image to be decomposed carries out LP conversion (Lars pyramid transform) with obtaining a series of band logical image under
Sample image.Assuming that A is input image, carrying out J LP to A converts to obtain a band logical image BjWith the image A after low-pass filteringj,
J=1,2,3 ... J (from detail to sparse alternation).And the LP of jth time is decomposed Aj-1It is divided into the image A after a down-samplingjWith
The band logical image B of one non-down-samplingj.When it is implemented, the value of J and the value of n are consistent.
2. the edge singular point that utilization orientation filtering group DFB will be distributed in same direction connects into coefficient.It is converted to LP
Band logical image B afterwardsjAfter carrying out sub-band division, the subband image in all directions is resolved into band logical subband image progress DFB.
Assuming that BjCarry out IjThe trend pass filtering of layer, then will obtainThe subband image in a directionWherein
Step 6, the low frequency for the reference low frequency image for removing the smooth image after foreground target and its corresponding region is utilized
Part carries out the low frequency part of individual raw video the low frequency after radiation adjustment is adjusted using differential technique or ratio method:
It is rightWithCarry out further low-pass filtering treatment, i=1,2,3...N, processing obtains corresponding more expressing
Image low frequencyWhen it is implemented, those skilled in the art, which can refer to the prior art, voluntarily specifies low-pass filtering
Handle realization method;Utilize the low-pass filtering result of the conversion low frequency part of acquisitionIt is and low after original conversion
FrequentlyResult after being adjusted using following difference or ratio formula
Or
Step 7, Contourlet inverse transformations are carried out using the low frequency part after the high frequency section of individual image and adjustment to obtain
Image ZI to after individual radiation adjustmenti, i=1,2,3...N:
It utilizesIt replacesAnd it combines originalCarry out the Contourlet inverse transformations even color processing that obtains that treated
Image ZI afterwardsi, i=1,2,3...N.When it is implemented, the contourlet transformation of Contourlet inverse transformations and step 5 is specific
It realizes accordingly, is its inverse process, it will not go into details by the present invention.
Step 8, to the image ZI after all N radiation adjustmentiSpliced to obtain final light and color homogenization result splicing shadow
As image after Z and final even color.
Referring to used in Fig. 2 and Fig. 3 present invention method tested obtained by image joint before and after light and color homogenization
Figure, it is seen then that this method takes the radiation two-dimensional distributed intelligence of image in itself into account, and first removes image and highlight or excessively dark target, subtracts
Lack the harmful effect to subsequently obtaining mean difference.Low-and high-frequency separation is carried out to image using contourlet transformation, in low frequency
Mean difference between the reference image that part statistical picture obtains carries out radiation consistency treatment.In processing procedure to image into
Row with overlapping region special piecemeal handle, only retain segmented areas " effective coverage " in as a result, the size of overlay region
It is codetermined by the low-pass filtering window size of the contourlet transformation number of plies and low frequency part.This method can also be applied and swept with pushing away
Due to Strip phenomenon caused by the difference in illumination and different shade directions in formula image.Can to different times image due to color and
Brightness is inconsistent to carry out good even color adjustment, and can preferable holding image detail information itself.
When it is implemented, method provided by the present invention, which can be based on software technology, realizes automatic running flow, mould can also be used
Block mode realizes corresponding system.
The present invention correspondingly provides the light and color homogenization system between a kind of image for taking radiation two-dimensional distribution into account, including with lower die
Block:
First module, for setting shared N image I to be splicedi, i=1,2,3...N, utilize the smoothing method of object level
Remove individual image I to be splicediIn highlighted or excessively dark foreground target, obtain object level it is smooth after image SIi, i=1,2,
3...N;
Second module, for count all object levels it is smooth after image SIiAverage SIimWith variance SIiv, obtain smooth
Whole average m and variance v afterwards;
3rd module, for according to average m and variance v, it is smooth to each object level after image SIiIt carries out being based on average side
The radiation adjustment of difference obtains radiation adjustment result image Ii', i=1,2,3...N;
4th module, for all N radiation adjustment result image Ii' carry out it is pre-splicing, and to pre-splicing result into
Row smoothly obtains whole with reference to low-frequency information image B;
5th module, for Ii、SIiAnd BIiContourlet transformation is carried out respectively obtains corresponding low frequency part And high frequency sectionWherein BIiIntegrally to be treated with reference to basis on low-frequency information image B
Splice image IiThe image that correspondence position is cut, i=1,2,3...N;
6th module, for pairWithFurther low-pass filtering treatment is carried out, obtains low-pass filtering resultWithUtilize the low-pass filtering result of the conversion low frequency part of acquisitionAndIt is public using difference or ratio
Formula be adjusted after resultI=1,2,3...N;
The difference value equation is as follows,
The ratio formula is as follows,
7th module, for utilizingIt replacesAnd combine original high frequency sectionCarry out Contourlet inversions
It changes, image ZI after the even color processing that obtains that treatedi, i=1,2,3...N;
8th module, for the image ZI after all N radiation adjustmentiSpliced to obtain final light and color homogenization result
Splice image Z.
Specific embodiment described herein is only to spirit explanation for example of the invention.Technology belonging to the present invention is led
The technical staff in domain can do various modifications or additions to described specific embodiment or replace in a similar way
Generation, but without departing from spirit of the invention or beyond the scope of the appended claims.
Claims (4)
1. a kind of light and color homogenization method between image for taking radiation two-dimensional distribution into account, which is characterized in that comprise the following steps:
Step 1, if shared N image I to be splicedi, i=1,2,3...N, it removes individual using the smoothing method of object level and waits to spell
Meet image IiIn highlighted or excessively dark foreground target, obtain object level it is smooth after image SIi, i=1,2,3...N;
Step 2, count all object levels it is smooth after image SIiAverage SIimWith variance SIiv, obtain the equal of smooth rear entirety
Value m and variance v;
Step 3, according to average m and variance v, it is smooth to each object level after image SIiCarry out the radiation tune based on mean variance
It is whole, obtain radiation adjustment result image Ii', i=1,2,3...N;
Step 4, to all N radiation adjustment result image Ii' progress is pre-splicing, and pre-splicing result is smoothly obtained whole
Body refers to low-frequency information image B;
Step 5, to Ii、SIiAnd BIiContourlet transformation is carried out respectively obtains corresponding low frequency partWith
High frequency sectionWherein BIiIntegrally to refer on low-frequency information image B according to image I to be splicediIt is corresponding
The image that position is cut, i=1,2,3...N;
Step 6, it is rightWithFurther low-pass filtering treatment is carried out, obtains low-pass filtering resultWithUtilize acquisition
Conversion low frequency part low-pass filtering resultAndIt is adjusted using difference value equation or ratio formula
Result afterwards
The difference value equation is as follows,
<mrow>
<msubsup>
<mi>I</mi>
<mi>i</mi>
<msup>
<mi>L</mi>
<mo>&prime;</mo>
</msup>
</msubsup>
<mo>=</mo>
<msubsup>
<mi>I</mi>
<mi>i</mi>
<mi>L</mi>
</msubsup>
<mo>-</mo>
<mrow>
<mo>(</mo>
<msubsup>
<mi>SI</mi>
<mi>i</mi>
<mrow>
<mi>L</mi>
<mi>L</mi>
</mrow>
</msubsup>
<mo>-</mo>
<msubsup>
<mi>BI</mi>
<mi>i</mi>
<mrow>
<mi>L</mi>
<mi>L</mi>
</mrow>
</msubsup>
<mo>)</mo>
</mrow>
<mo>=</mo>
<msubsup>
<mi>I</mi>
<mi>i</mi>
<mi>L</mi>
</msubsup>
<mo>-</mo>
<msubsup>
<mi>SI</mi>
<mi>i</mi>
<mrow>
<mi>L</mi>
<mi>L</mi>
</mrow>
</msubsup>
<mo>+</mo>
<msubsup>
<mi>BI</mi>
<mi>i</mi>
<mrow>
<mi>L</mi>
<mi>L</mi>
</mrow>
</msubsup>
</mrow>
The ratio formula is as follows,
<mrow>
<msubsup>
<mi>I</mi>
<mi>i</mi>
<msup>
<mi>L</mi>
<mo>&prime;</mo>
</msup>
</msubsup>
<mo>=</mo>
<msubsup>
<mi>I</mi>
<mi>i</mi>
<mi>L</mi>
</msubsup>
<mo>/</mo>
<mrow>
<mo>(</mo>
<msubsup>
<mi>SI</mi>
<mi>i</mi>
<mrow>
<mi>L</mi>
<mi>L</mi>
</mrow>
</msubsup>
<mo>/</mo>
<msubsup>
<mi>BI</mi>
<mi>i</mi>
<mrow>
<mi>L</mi>
<mi>L</mi>
</mrow>
</msubsup>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mrow>
<mo>(</mo>
<msubsup>
<mi>I</mi>
<mi>i</mi>
<mi>L</mi>
</msubsup>
<mo>/</mo>
<msubsup>
<mi>SI</mi>
<mi>i</mi>
<mrow>
<mi>L</mi>
<mi>L</mi>
</mrow>
</msubsup>
<mo>)</mo>
</mrow>
<mo>*</mo>
<msubsup>
<mi>BI</mi>
<mi>i</mi>
<mrow>
<mi>L</mi>
<mi>L</mi>
</mrow>
</msubsup>
</mrow>
Step 7, utilizeIt replacesAnd combine original high frequency sectionContourlet inverse transformations are carried out, after obtaining processing
Even color processing after image ZIi, i=1,2,3...N;
Step 8, to the image ZI after all N radiation adjustmentiSpliced to obtain final light and color homogenization result splicing image Z.
2. a kind of light and color homogenization method between image for taking radiation two-dimensional distribution into account according to claim 1, it is characterised in that:
In step 3, the radiation adjustment realization method based on mean variance is as follows,
IfFor SIiKth wave band SIikAverage gray,It is the average for the whole kth wave band that statistics obtains,With
It is the standard deviation of corresponding wave band, mean variance method radiant correction coefficient is:
<mrow>
<msub>
<mi>a</mi>
<mi>k</mi>
</msub>
<mo>=</mo>
<mfrac>
<msub>
<mi>S</mi>
<msub>
<mi>y</mi>
<mi>k</mi>
</msub>
</msub>
<msub>
<mi>S</mi>
<msub>
<mi>x</mi>
<mi>k</mi>
</msub>
</msub>
</mfrac>
</mrow>
<mrow>
<msub>
<mi>b</mi>
<mi>k</mi>
</msub>
<mo>=</mo>
<mover>
<msub>
<mi>y</mi>
<mi>k</mi>
</msub>
<mo>&OverBar;</mo>
</mover>
<mo>-</mo>
<msub>
<mi>a</mi>
<mi>k</mi>
</msub>
<mo>&CenterDot;</mo>
<mover>
<msub>
<mi>x</mi>
<mi>k</mi>
</msub>
<mo>&OverBar;</mo>
</mover>
</mrow>
Then I 'iKth wave band be I 'ik=akSIik+bk。
3. the light and color homogenization system between a kind of image for taking radiation two-dimensional distribution into account, which is characterized in that including with lower module:
First module, for setting shared N image I to be splicedi, i=1,2,3...N, it is removed using the smoothing method of object level single
Open image I to be splicediIn highlighted or excessively dark foreground target, obtain object level it is smooth after image SIi, i=1,2,3...N;
Second module, for count all object levels it is smooth after image SIiAverage SIimWith variance SIiv, obtain smooth rear whole
The average m of body and variance v;
3rd module, for according to average m and variance v, it is smooth to each object level after image SIiIt carries out based on mean variance
Radiation adjustment obtains radiation adjustment result image Ii', i=1,2,3...N;
4th module, for all N radiation adjustment result image Ii' progress is pre-splicing, and pre-splicing result is carried out smooth
It obtains whole with reference to low-frequency information image B;
5th module, for Ii、SIiAnd BIiContourlet transformation is carried out respectively obtains corresponding low frequency part And high frequency sectionWherein BIiIntegrally to refer on low-frequency information image B according to image I to be splicedi
The image that correspondence position is cut, i=1,2,3...N;
6th module, for pairWithFurther low-pass filtering treatment is carried out, obtains low-pass filtering resultWith
Utilize the low-pass filtering result of the conversion low frequency part of acquisitionAndUtilize difference value equation or ratio formula
Result after being adjusted
The difference value equation is as follows,
<mrow>
<msubsup>
<mi>I</mi>
<mi>i</mi>
<msup>
<mi>L</mi>
<mo>&prime;</mo>
</msup>
</msubsup>
<mo>=</mo>
<msubsup>
<mi>I</mi>
<mi>i</mi>
<mi>L</mi>
</msubsup>
<mo>-</mo>
<mrow>
<mo>(</mo>
<msubsup>
<mi>SI</mi>
<mi>i</mi>
<mrow>
<mi>L</mi>
<mi>L</mi>
</mrow>
</msubsup>
<mo>-</mo>
<msubsup>
<mi>BI</mi>
<mi>i</mi>
<mrow>
<mi>L</mi>
<mi>L</mi>
</mrow>
</msubsup>
<mo>)</mo>
</mrow>
<mo>=</mo>
<msubsup>
<mi>I</mi>
<mi>i</mi>
<mi>L</mi>
</msubsup>
<mo>-</mo>
<msubsup>
<mi>SI</mi>
<mi>i</mi>
<mrow>
<mi>L</mi>
<mi>L</mi>
</mrow>
</msubsup>
<mo>+</mo>
<msubsup>
<mi>BI</mi>
<mi>i</mi>
<mrow>
<mi>L</mi>
<mi>L</mi>
</mrow>
</msubsup>
</mrow>
The ratio formula is as follows,
<mrow>
<msubsup>
<mi>I</mi>
<mi>i</mi>
<msup>
<mi>L</mi>
<mo>&prime;</mo>
</msup>
</msubsup>
<mo>=</mo>
<msubsup>
<mi>I</mi>
<mi>i</mi>
<mi>L</mi>
</msubsup>
<mo>/</mo>
<mrow>
<mo>(</mo>
<msubsup>
<mi>SI</mi>
<mi>i</mi>
<mrow>
<mi>L</mi>
<mi>L</mi>
</mrow>
</msubsup>
<mo>/</mo>
<msubsup>
<mi>BI</mi>
<mi>i</mi>
<mrow>
<mi>L</mi>
<mi>L</mi>
</mrow>
</msubsup>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mrow>
<mo>(</mo>
<msubsup>
<mi>I</mi>
<mi>i</mi>
<mi>L</mi>
</msubsup>
<mo>/</mo>
<msubsup>
<mi>SI</mi>
<mi>i</mi>
<mrow>
<mi>L</mi>
<mi>L</mi>
</mrow>
</msubsup>
<mo>)</mo>
</mrow>
<mo>*</mo>
<msubsup>
<mi>BI</mi>
<mi>i</mi>
<mrow>
<mi>L</mi>
<mi>L</mi>
</mrow>
</msubsup>
</mrow>
7th module, for utilizingIt replacesAnd combine original high frequency sectionContourlet inverse transformations are carried out, are obtained
Image ZI after even color processing that treatedi, i=1,2,3...N;
8th module, for the image ZI after all N radiation adjustmentiSpliced to obtain final light and color homogenization result splicing
Image Z.
4. a kind of light and color homogenization system between image for taking radiation two-dimensional distribution into account according to claim 3, it is characterised in that:
In 3rd module, the radiation adjustment realization method based on mean variance is as follows,
IfFor SIiKth wave band SIikAverage gray,It is the average for the whole kth wave band that statistics obtains,With
It is the standard deviation of corresponding wave band, mean variance method radiant correction coefficient is:
<mrow>
<msub>
<mi>a</mi>
<mi>k</mi>
</msub>
<mo>=</mo>
<mfrac>
<msub>
<mi>S</mi>
<msub>
<mi>y</mi>
<mi>k</mi>
</msub>
</msub>
<msub>
<mi>S</mi>
<msub>
<mi>x</mi>
<mi>k</mi>
</msub>
</msub>
</mfrac>
</mrow>
<mrow>
<msub>
<mi>b</mi>
<mi>k</mi>
</msub>
<mo>=</mo>
<mover>
<msub>
<mi>y</mi>
<mi>k</mi>
</msub>
<mo>&OverBar;</mo>
</mover>
<mo>-</mo>
<msub>
<mi>a</mi>
<mi>k</mi>
</msub>
<mo>&CenterDot;</mo>
<mover>
<msub>
<mi>x</mi>
<mi>k</mi>
</msub>
<mo>&OverBar;</mo>
</mover>
</mrow>
Then I 'iKth wave band be I 'ik=akSIik+bk。
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610111326.0A CN105787896B (en) | 2016-02-29 | 2016-02-29 | A kind of light and color homogenization method and system between the image for taking radiation two-dimensional distribution into account |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610111326.0A CN105787896B (en) | 2016-02-29 | 2016-02-29 | A kind of light and color homogenization method and system between the image for taking radiation two-dimensional distribution into account |
Publications (2)
Publication Number | Publication Date |
---|---|
CN105787896A CN105787896A (en) | 2016-07-20 |
CN105787896B true CN105787896B (en) | 2018-06-01 |
Family
ID=56403169
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610111326.0A Active CN105787896B (en) | 2016-02-29 | 2016-02-29 | A kind of light and color homogenization method and system between the image for taking radiation two-dimensional distribution into account |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105787896B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106910235A (en) * | 2017-01-20 | 2017-06-30 | 同济大学 | Territorial laser scanning data light and color homogenization method based on Coordinate Conversion |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103489171A (en) * | 2013-09-22 | 2014-01-01 | 武汉大学 | Wide-range remote-sensing image automatic dodging and color uniformizing method based on standard color library |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR102054680B1 (en) * | 2013-01-23 | 2020-01-22 | 삼성전자주식회사 | Image processing apparatus, ultrasonic imaging apparatus and method for image processing |
-
2016
- 2016-02-29 CN CN201610111326.0A patent/CN105787896B/en active Active
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103489171A (en) * | 2013-09-22 | 2014-01-01 | 武汉大学 | Wide-range remote-sensing image automatic dodging and color uniformizing method based on standard color library |
Non-Patent Citations (3)
Title |
---|
Algorithm for relative radiometric consistency process of remote sensing images based on object-oriented smoothing and contourlet transforms;Wenzhuo Li et al.;《Journal of Applied Remote Sensing》;20140625;第8卷(第1期);第1-13页 * |
Automatic radiometric normalization of multitemporal satellite imagery with the iteratively re-weighted MAD transformation;Morton J. Canty et al.;《Remote Sensing of Environment》;20080318;第112卷(第3期);第1025-1036页 * |
基于INPHO在建库影像匀光匀色中的应用;孙冬梅;《测绘技术装备》;20101231;第12卷(第4期);第21-23页 * |
Also Published As
Publication number | Publication date |
---|---|
CN105787896A (en) | 2016-07-20 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US10102614B2 (en) | Fog removing device and image generating method | |
CN103955905B (en) | Based on the single image to the fog method that fast wavelet transform and weighted image merge | |
CN110175964A (en) | A kind of Retinex image enchancing method based on laplacian pyramid | |
CN103034986A (en) | Night vision image enhancement method based on exposure fusion | |
CN107358585B (en) | Foggy day image enhancement method based on fractional order differential and dark channel prior | |
CN107240081A (en) | The denoising of night scene image and enhancing processing method | |
Wang et al. | Variational single nighttime image haze removal with a gray haze-line prior | |
CN109215042B (en) | Photovoltaic cell panel hot spot effect detection system based on computer vision and calculation method thereof | |
Pei et al. | Effective image haze removal using dark channel prior and post-processing | |
CN110163807B (en) | Low-illumination image enhancement method based on expected bright channel | |
CN106204494A (en) | A kind of image defogging method comprising large area sky areas and system | |
CN106651817A (en) | Non-sampling contourlet-based image enhancement method | |
CN102646267B (en) | Degraded image restoration method and system | |
Iqbal et al. | Color and white balancing in low-light image enhancement | |
CN106971380A (en) | A kind of contrast enhancing and application of the visual saliency optimization method in golf course figure | |
Kumar et al. | Enhancing scene perception using a multispectral fusion of visible–near‐infrared image pair | |
Zhu et al. | Low-light image enhancement network with decomposition and adaptive information fusion | |
CN105787896B (en) | A kind of light and color homogenization method and system between the image for taking radiation two-dimensional distribution into account | |
CN110175959B (en) | Typhoon cloud picture enhancement method | |
CN110111268B (en) | Single image rain removing method and device based on dark channel and fuzzy width learning | |
Mathur et al. | Dual domain approach for colour enhancement of underwater images | |
Tang et al. | Sky-preserved image dehazing and enhancement for outdoor scenes | |
CN109886901B (en) | Night image enhancement method based on multi-channel decomposition | |
Yehia et al. | An improved integrated intensity-hue-saturation with stationary wavelet transform multi-sensor image fusion approach | |
Kaur et al. | Image enhancement of underwater digital images by utilizing L* A* B* color space on gradient and CLAHE based smoothing |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |