Multiscale Retinex Technique with

1
Multiscale Retinex Technique with Color
Correction Based on the Retinex Theory
2
Retinex Image Enhancement

• Edward Land86
• There exists a discrepancy between the human
  vision system and the recorded color images.
• Dynamic range difference results in the loss of
  the essentials features from the recorded images.
• Improved fidelity of color images to human
  observation can be obtained by
• (a) Computation that combines dynamic range
• compression, color constancy and color
  rendition
• (b) Color restoration.

3
Block Diagram

I(x,y)
MSR
Log
W1

F1(x,y)
-
W2


Gain/ offset

Log

S
S
MSRCR
F2(x,y)
W3


F3(x,y)
a
CRF
CR
Log
4
Design of Surround Function

• First proposed design of Surround function by
  E.Land86 was inverse square spatial surround
• F(x,y) 1/ 1(r2 c2)
• The surround function was later modified in
  Gaussian form
• by Hurlbert89
• F(x,y) exp(-r2 / c2 )
• Where
• r- v x2 y2 and
• c- Surround Space Constant

5
Single scale Retinex Method

• The Single Scale retinex is given by
• Ri (x,y)log Ii (x,y) log F(x,y) Ii (x,y)
• Where
• F(x,y) K exp(-r2 / c2 )--- Surround Function
• c- Scalar value and selection of K is that
• r- v x2 y2
• ?? F(x,y) dx dy 1

6
Multiscale Retinex Method

• The multi-scale retinex is represented by
• Ri (x,y) S Wn log Ii (x,y) - log F(x,y)
  Ii (x,y)
• Where
• n -- Scaling Factor
• Wn Weights (1/3 for each color channel of RGB)

N
n1

http//dragon.larc.nasa.gov/pub/papers/multsclrtx.
pdf
7
Limitations of MSR

• Limitations of the MSR
• The Selection of the value of c in equ(1) is
  critical.
• The DRC results in the violation of Gray world
  algorithm
• The region of constant color bleaches out as a
  result of DRC.
• Gray World Assumptions
• Gray World Assumption states is that, given an
  image with sufficient
• amount of color variations, the average value of
  the RED, GREEN, and
• BLUE components of the image should average out
  to a common gray
• value.

8
Color Restoration

• The color restoration is calculated using the
  expression
• Ci(x,y) ßloga Ii (x,y) log S Ii(x,y)
• Where
• ß- Gain Constant
• a- Controls the strength of non-linearity
• The Final representation of MSRCR is represented
  as
• R MSRCRi (x, y) G Ci (x, y) RMSRi
  (x, y) b
• Where
• G- Gain Constant and
• b- Gain Offset value

s
i1
9
ALGORITHM-I
RESULTS
10
Constants Values Applied in Retinex

• Wn - 1/3 - Weight used in Multiscale Retinex
• N Number of Scale 3
• C1, C2, C3 - Surround Constant 15, 80,250
  respectively
• G - Final Gain 192
• b - Offset Value 30
• a Strength of non-linearity 125
• ß Control gain constant - 46

11
Surround Function
Gaussian Surround Function
F(r)
Image Co-ordinate
12
Results of SSR

Space Constant c80
Input
Output
13
Results of SSR
Space Constant c80
Inputs
Outputs
14
Results of SSR
Outputs
Inputs
c15
c80
c215
15
Results of MSRCR
Output of MSRCR
Inputs
MSR Output
MSRCR Output
16
Results of MSRCR

• Tak at IRIS Laboratory

Inputs
MSR Output
MSRCR Output
17
Results of MSRCR

• Tak at McGhee Tyson Airport

Inputs
MSR Output
MSRCR Output
18
Comparison- Available Software and My
Implementation

• Few Examples
• Input Software My results

19
Comparison- Available Software and My
Implementation

• Tak at IRIS and McGhee Tyson Airport
• Input
  Software My results

20
Comparison- Available Software and My
Implementation

• The following image was presented as an example
  in the paper, the same
• image is used as input to both the software
  available and my
• implementation.

My Implementation
From Available Software
Test Image
http//dragon.larc.nasa.gov/pub/papers/multsclrtx.
pdf