Color Transfer between Images

1
Color Transfer between Images

• ReporterChen Shuangmin

2
Introduction to the authors

• Erik Reinhard
• .is a postdoctoral fellow at the University of
  Utah
• .interested in visual-perception and
  parllelgraphics fields.
• .BS diploma in computer science from Deft
• University of Technology and a phD in
  computer science from the university of
  Bristol

3
Introduction to the authors

• Bruce Gooch
• .is a graduate student at the university of
  Utah
• .intersted in nonphotorealistic rendering
• .BS in mathematics and an MS in computer
  science from the university of Utah

4
Introduction to the authors

• Michael Ashikhmin
• .an assistant professor at State University of
  New York
• .intersted in developing simple and practical
  algorithms for computer graphics
• .has an MS in physics from the Moscow
  Institute of Physics and Technology,an MS
• in chemistry from the University of California

5
Guidence

• Color correction(remove undesired
  color-illumination) - in one image

Red eye restoration
6
Guidence

• Color correction(remove undesired
  color-illumination)
• Borrow color from one image to another

source
target
Similar composition
7
Main idea

• 1.Color constancy correction
• .global statistics
• .swatches
• 2.Gamma correction(logxr rlogx)
• Goal
• Alter an images color

8
1.Color constancy correction
--- Gloabal statistics

• Why choose lab color space???
• .RGB
• .lab
• .CIECAM97s
• Why decorrelation is better???

9
Introduction to RGB

• We can get any color by blending with RGB

Definition with cubic
10
Introduction to lab color space

• In Photoshop,the range of l is 0 to 100and in
  color palete,the ranges of a and b are both 120
  to -120.

It will be more easily understood in Gamma
correction
11
Why choose lab color space?
--decorrelation
2000 random points
12
Why decorrelation is better?

• Apply different operations in different color
  channels that without undesirable crosschannel
  artifacts.
• The data in lab space is more compressed
  (log space)
• RGBcorrelations between the different channels
  values. (0,0,0) - (255,255,255)

13
Comparison of the effects
In three color spaces
Source image
Target image
14
Comparison of effects
In three color spaces
RGB color space
CIECAM97s
Lab color space
15
Convertion between color spaces

• RGB lab
• .RGB XYZ (international standard)
• .XYZ LMS
• .LMS LMS
• .LMS lab
• lab RGB

16
About CIECAM97s

• RGB XYZ (international standard)
• XYZ LMS

17
Convertion between color spaces

• RGB lab
• .RGB XYZ
• .XYZ LMS

18
Convertion between color spaces

• RGB lab
• .LMS LMS
• .LMS lab

19
Convertion between color spaces

• RGB lab
• lab RGB(inverse operations)

L 10L M 10M S 10S
20
Main idea

• 1.Color constancy correction
• .global statistics
• .swatches
• 2.Gamma correction(logxr rlogx)

21
Two concept in statistics

• Standard deviation
• Mean value

Show with chalk on blackboard
22
Color constancy correction

• Subtract the mean from the data points
• Scale the data points comprising the synthetic
  image by factors determined by the respective
  standard deviation

23
Color constancy correction

• add the averages computed for the photagraph.
• Displaying convert the result back to RGB via
  logLMS,LMS,and XYZ color spaces using equations

ltltgt
ltatgt
ltbtgt
24
Convert back to RGB for displaying

• RGB lab
• lab RGB(inverse operations)

L 10L M 10M S 10S
25
Main idea

• 1.Color constancy correction
• .global statistics
• .swatches
• 2.Gamma correction(logxr rlogx)

26
Swatches

• When source and target images dont work well
  together swatches.

27
Swatches

• Several pairs of clusters in lab space (cluster
  for each swatch)
• Compute statistics for each swatch

28
Swatches -result

• When source and target images dont work well
  swatches.

target
source
result
29
Main idea

• 1.Color constancy correction
• .global statistics
• .swatches
• 2.Gamma correction(logxr r logx)

30
My question about this paper

• I dont know how to decide which is source image
  and which is target image. It will give different
  effects.

T
result
T
S
result
S
This effect is much better! I like it.
I dont like this result!
31
My question about this paper

• I dont know which two images will be matching
  better .How to choose?

result
result
This effect is much better! I like it.(result of
this paper.)
This effect is not much better!
32
My question about this paper

• The size of these two images is related to the
  result.

Target 800600 Source 426266
Target 454366 Source 426266
33

• Thank you!