EECE253 03 PointProcessing

EECE/CS 253 Image Processing
Lecture Notes: Lecture Notes: The Point Processing of Images Richard Alan Peters II Department of Electrical Engineering and Computer Science Fall Semester 2011
This work is licensed under the Creative Commons Attribution-Noncommercial 2.5 License. To view a copy of this license, visit http://creativecommons.org/licenses/by-nc/2.5/ or send a letter to Creative Commons, 543 Howard Street, 5th Floor, San Francisco, California, 94105, USA.
Point Processing of Images

In a digital image, point = pixel. Point processing transforms a pixels value as function of its value alone; it does not depend on the values of the pixels neighbors.
2011-09-06
2000-2011 Richard Alan Peters II
Point Processing of Images

Brightness and contrast adjustment Gamma correction Histogram equalization Histogram matching Color correction.
2011-09-06
Point Processing
- gamma
- brightness
original
+ brightness
+ gamma
histogram mod
2011-09-06
- contrast
original
+ contrast
histogram EQ
4
The Histogram of a Grayscale Image

Let I be a 1-band (grayscale) image. I(r,c) is an 8-bit integer between 0 and 255. Histogram, hI, of I:

a 256-element array, hI hI(g), for g = 1, 2, 3, , 256, is an integer hI(g) = number of pixels in I that have value g-1.
2011-09-06
lower RHC: number of pixels with intensity g
black marks pixels with intensity g
16-level (4-bit) image
2011-09-06

Black marks pixels with intensity g
Plot of histogram: number of pixels with intensity g
2011-09-06

Black marks pixels with intensity g
Plot of histogram: number of pixels with intensity g
2011-09-06
Luminosity
2011-09-06
The Histogram of a Color Image

If I is a 3-band image (truecolor, 24-bit) then I(r,c,b) is an integer between 0 and 255. Either I has 3 histograms:

hR( g +1) = # of pixels in I(:,:,1) with intensity value g hG( g +1) = # of pixels in I(:,:,2) with intensity value g hB( g +1) = # of pixels in I(:,:,3) with intensity value g h( g +1,1,1) = # of pixels in I with red intensity value g h( g +1,1,2) = # of pixels in I with green intensity value g h( g +1,1,3) = # of pixels in I with blue intensity value g
or 1 vector-valued histogram, h(g,1,b) where

2011-09-06
10

There is one histogram per color band R, G, & B. Luminosity histogram is from 1 band = (R+G+B)/3
Luminosity
2011-09-06
11
2011-09-06
12
Value or Luminance Histograms

The value histogram of a 3-band (truecolor) image, I, is the histogram of the value image,
1 V (r , c) = [ R (r , c) + G (r , c) + B (r , c)] 3
Where R, G, and B are the red, green, and blue bands of I. The luminance histogram of I is the histogram of the luminance image,
L (r , c) = 0.299 R (r , c) + 0.587 G (r , c) + 0.114 B (r , c)
2011-09-06
13
Value Histogram
Value image, V.
Histogram of the value image.
2011-09-06
14
Luminance Histogram
Luminance image, L.
Histogram of the luminance image.
2011-09-06
15
Value Histogram vs. Average of R,G,&B Histograms
R,G,B,&V histograms
2011-09-06
V & avg. of R,G,&B histograms

16
Multi-Band Histogram Calculator in Matlab

% Multi-band histogram calculator function h=histogram(I) [R C B]=size(I); % allocate the histogram h=zeros(256,1,B); % range through the intensity values for g=0:255 h(g+1,1,:) = sum(sum(I==g)); % accumulate end return;
2011-09-06
17
Multi-Band Histogram Calculator in Matlab

% Multi-band histogram calculator Loop through all intensity levels (0-255) function h=histogram(I) Tag the elements that have value g.
The result is an RxCxB logical array that [R C B]=size(I); has a 1 wherever I(r,c,b) = g and 0s everywhere else. Compute the number of ones in each band of % allocate the histogram the image for intensity g. h=zeros(256,1,B); Store that value in the 256x1xB histogram at h (g+1,1,b). % range through the intensity values
for g=0:255 h(g+1,1,:) = sum(sum(I==g)); % accumulate end

If B==3, then h(g+1,1,:) contains return; 3 numbers: the number of pixels in bands 1, 2, & 3 that have intensity g.
2011-09-06
sum(sum(I==g)) computes one number for each band in the image.
18
Point Ops via Functional Mappings

Image:
I
Input
, point
operator
J
Output
J = F[ I ]
Pixel:
I(r,c)
function, f
J(r,c)
If I (r , c) = g and f ( g ) = k
The transformation of image I into image J is then J (r , c) = k . accomplished by replacing each input intensity, g, with a specific output intensity, k, at every location (r,c) where I(r,c) = g. The rule that associates k with g is usually specified with a function, f, so that f (g) = k.
2011-09-06
19

One-band Image
Three-band Image
2011-09-06
20

One-band Image
Either all 3 bands are mapped through the same function, f, or
Three-band Image
each band is mapped through a separate function, fb.

2011-09-06
21
Point Operations using Look-up Tables

A look-up table (LUT) implements a functional mapping.
If k = f ( g ), for g = 0,, 255,
then the LUT that implements f is a 256x1 array whose (g +1)th value is k = f (g).
and if k takes on values in {0,, 255},

2011-09-06
To remap a 1-band image, I, to J :
22
Point Operations using Look-up Tables

If I is 3-band, then a) each band is mapped separately using the same LUT for each band or b) each band is mapped using different LUTs one for each band.
2011-09-06
23
Point Operations = Look-up Table Ops

E.g.: index value ... ... 64 101 68 102 103 69 70 104 105 70 106 71 ... ... input output
24
output value 127
255
127 input value
255
2011-09-06
Look-Up Tables
input cell index
a pixel with this value
contents
0 . . . 64 . . . 128 . . . 192 255
. . .
0 . . . 32 . . . 128 . . . 224 255
output
is mapped to this value
. . .
2011-09-06
25
How to Generate a Look-Up Table

For example:
Let a = 2. Let x {0,,255}
(x ; a ) =
Or in Matlab:
255 1 + e a ( x 127 ) / 32
a = 2; x = 0:255; LUT = 255 ./ (1+exp(-a*(x-127)/32));
This is just one example.

26
2011-09-06
Point Processes: Original Image
Kinkaku-ji ( Temple of the Golden Pavilion), also known as Rokuon-ji ( Deer Garden Temple), is a Zen Buddhist temple in Kyoto, Japan.
Photo by Richard Alan Peters II, August 1993.
Luminance Histogram
2011-09-06
27
Point Processes: Increase Brightness
J ( r , c , b) =
I (r , c, b) + g , if I (r , c, b) + g < 256 if I (r , c, b) + g > 255 255,
127
255
g
saturation point
0 0 127 255
g 0 and b { 1, 2,3 } is the band index.
transform mapping
28
2011-09-06
Point Processes: Decrease Brightness
255
zero point
255-g
J ( r , c , b) =
if I (r , c, b) - g < 0 0, I (r , c, b) - g , if I (r , c, b) - g > 0
0 0 127 255
g 0 and b { 1, 2,3 } is the band index.
127
transform mapping
29
2011-09-06
Point Processes: Decrease Contrast
Here, s = 127
T( r , c , b ) = a [ I ( r , c , b) - s ] + s ,
s is the where 0 a < 1.0, center of s {0,1, 2, , 255} , and the contrast function. b { 1, 2,3 } .
127 0
255
127
255
transform mapping
30
2011-09-06
Point Processes: Increase Contrast
255
zero point
Here, s = 127
T ( r , c , b ) = a [ I ( r , c , b) - s ] + s
0, J (r , c, b) = T ( r , c, b) , 255, a > 1, s {0,, 255} ,
2011-09-06
127
if T (r , c, b) < 0, if 0 T (r , c, b) 255, if T (r , c, b) > 255.
sat. point
0 0 127 255
b { 1, 2 ,3 }
transform mapping
31
Point Processes: Contrast Stretch
MJ
255 127
zero point
Let mI = min [I (r , c)] , M I = max [I (r , c)], mJ = min [ J (r , c)] , M J = max [ J (r , c)]. Then, J (r , c) = ( M J - mJ )
center point
I (r , c) - mI + mJ . M I - mI
mJ
sat. point
0 127 mI MI 255 transform mapping
2011-09-06
32
histograms
Information Loss from Contrast Adjustment

orig
lo-c
hi-c
2011-09-06
33
Information Loss from Contrast Adjustment

orig lo-c hi-c
abbreviations: original low-contrast high-contrast restored difference
orig
lo-c rest
lo-c diff hi-c diff

2000-2011 Richard Alan Peters II difference between original and restored high-contrast difference between original and restored low-contrast
orig
hi-c rest
2011-09-06
34
Point Processes: Increased Gamma
I ( r , c) g J ( r , c) = 255 for g > 1.0 255
127 0
255
127
255
transform mapping
35
2011-09-06
Point Processes: Decreased Gamma
I ( r , c) g J ( r , c) = 255 for g < 1.0 255

2011-09-06
127 0 127 m M 255 transform mapping
255
36
Gamma Correction: Effect on Histogram
2011-09-06
37
The Probability Density Function of an Image

Let A = hI k ( g + 1) .
g =0 255
pdf [lower case]
Note that since hI k ( g + 1) is the number of pixels in I k (the kth color band of image I ) with value g , A is the number of pixels in I. That is if I is R rows by C columns then A = R C.
Then, pIk ( g + 1) = 1 hIk ( g + 1) A
This is the probability that an arbitrary pixel from Ik has value g.
is the graylevel probability density function of I k .

2011-09-06
38
The Probability Density Function of an Image

pband(g+1) is the fraction of pixels in (a specific band of) an image that have intensity value g. pband(g+1) is the probability that a pixel randomly selected from the given band has intensity value g. Whereas the sum of the histogram hband(g+1) over all g from 1 to 256 is equal to the number of pixels in the image, the sum of pband(g+1) over all g is 1. pband is the normalized histogram of the band.
2011-09-06
39
The Probability Distribution Function of an Image

Let q = [q1 q2 q3] = I(r,c) be the value of a randomly selected pixel from I. Let g be a specific graylevel. The probability that qk g is given by
1 g PI k ( g + 1) = pIk (g + 1) = hI (g +1) = A g =0 k g =0
g
PDF [upper case]

g
h h
g =0 g =0 255
Ik
(g + 1)
,
Ik
(g + 1)
where hIk( +1) is the histogram of the kth band of I.
This is the probability that any given pixel from Ik has value less than or equal to g.
2011-09-06
40

Let q = [q1 q2 q3] = I(r,c) be the value of a randomly selected pixel from I. Let g be a specific graylevel. The probability that qk g is given by
1 g PI k ( g + 1) = pIk (g + 1) = hI (g +1) = A g =0 k g =0
g
Also called CDF for Cumulative Distribution Function.
h h
g =0 g =0 255
Ik
(g + 1)
,
Ik
(g + 1)
where hIk( +1) is the histogram of the kth band of I.
This is the probability that any given pixel from Ik has value less than or equal to g.
2011-09-06
41
A.k.a. Cumulative Distribution Function.

Pband(g+1) is the fraction of pixels in (a specific band of) an image that have intensity values less than or equal to g. Pband(g+1) is the probability that a pixel randomly selected from the given band has an intensity value less than or equal to g. Pband(g+1) is the cumulative (or running) sum of pband(g+1) from 0 through g inclusive. Pband(1) = pband(1) and Pband(256) = 1; Pband(g+1) is nondecreasing.
Note: the Probability Distribution Function (PDF, capital letters) and the Cumulative Distribution Function (CDF) are exactly the same things. Both PDF and CDF will refer to it. However, pdf (small letters) is the density function. 2011-09-06
42
Point Processes: Histogram Equalization

Task: remap image I so that its histogram is as close to constant as possible
Let PI (g + 1) be the cumulative (probability) distribution
function of I.
Then J has, as closely as possible, the correct histogram if
J (r , c, b) = 255 PI [I (r , c, b) + 1].
The CDF itself is used as the LUT.
all bands processed similarly

43
2011-09-06

Luminosity
before
J (r , c, b) = 255 PI ( g + 1) , g = I (r , c, b) , b {1, 2,3} .

2011-09-06
after
44
Histogram EQ
The CDF (cummulative distribution) is the pdf LUT for remapping.
CDF
2011-09-06
45
Histogram EQ
LUT
2011-09-06
46
Histogram EQ
LUT
2011-09-06
47
Histogram EQ
2011-09-06
48

Task: remap image I with min = mI and max = MI so that its histogram is as close to constant as possible and has min = mJ and max = MJ .
Let PI (g + 1) be the cumulative (probability) distribution
function of I.
Then J has, as closely as possible, the correct histogram if

Using intensity extrema
2011-09-06
PI [I (r , c) + 1]- PI (mI + 1) J (r , c) = ( M J - mJ ) + mJ . PI ( M I + 1) - PI (mI + 1)

49
Point Processes: Histogram Matching

Task: remap image I so that it has, as closely as possible, the same histogram as image J. Because the images are digital it is not, in general, possible to make hI hJ . Therefore, pI pJ . Q: How, then, can the matching be done? A: By matching percentiles.
2011-09-06
50
Matching Percentiles
Recall: The CDF of image I is such that 0 PI ( gI ) 1. PI ( gI +1) = c means that c is the fraction of pixels in I that have a value less than or equal to gI . 100c is the percentile of pixels in I that are less than or equal to gI . To match percentiles, replace all occurrences of value gI in image I with the value, gJ, from image J whose percentile in J most closely matches the percentile of gI in image I.
2011-09-06
assuming a 1-band image or a single band of a color image.
51
Matching Percentiles
So, to create an image, K, from image I such that K has nearly the same CDF as image J do the following: If I(r,c) = gI then let K(r,c) = gJ where gJ is such that
PI (gI) > PJ (gJ -1) AND PI (gI) PJ (gJ).

PI ( gI ) PJ ( gJ )
Example: I(r,c) = 5 PI (5) = 0.65 PJ (9) = 0.56 PJ (10) = 0.67 K(r,c) = 10
gI
2011-09-06
gJ
52
Histogram Matching Algorithm

[R,C] = size(I); K = zeros(R,C); gJ = mJ; for gI = mI to MI while g J < 255 AND PI ( gI + 1) < 1 AND PJ ( g J + 1) < PI ( g I + 1)
g J = g J + 1;
This directly matches image I to image J.

PJ ( g J + 1) : CDF of J. mJ = min J, PI ( g I + 1) : CDF of I
M J = max J, mI = min I, M I = max I.
end K = K + [ g J ( I == g I )] end
Better to use a LUT. See slide 58.

53
2011-09-06
Example: Histogram Matching

Image pdf
Image with 16 intensity values
2011-09-06
54

Image CDF
CDFI(g)
g
*a.k.a Cumulative Distribution Function, CDFI. 2011-09-06
55

Target pdf
Target with 16 intensity values
2011-09-06
56

Target CDF
CDFI(g)
g
*a.k.a Cumulative Distribution Function, CDFJ. 2011-09-06
57
Histogram Matching with a Lookup Table

The algorithm on slide 53 matches one image to another directly. Often it is faster or more versatile to use a lookup table (LUT). Rather than remapping each pixel in the image separately, one can create a table that indicates to which target value each input value should be mapped. Then
K = LUT[I+1]
In Matlab if the LUT is a 256 1 matrix with values from 0 to 255 and if image I is of type uint8, it can be remapped with the following code:
K = uint8(LUT(I+1));
2011-09-06
58
LUT Creation
Image CDF Target CDF
LUT
10
2011-09-06
59
Look Up Table for Histogram Matching

LUT = zeros(256,1); gJ = 0;
This creates a look-up table which can then be used to remap the image.
for gI = 0 to 255 while PJ ( g J + 1) < PI ( g I + 1) AND g J < 255 g J = g J + 1; end LUT ( g I + 1) = g J ; end
2011-09-06
60
Input & Target CDFs, LUT and Resultant CDF

Input
PJ ( g ) PI ( g )
Target
LUT ( g )
g
2011-09-06
PK ( g )
LUT
Result
g
61
original
2011-09-06
target
remapped
62
Probability Density Functions of a Color Image
Atlas-Mercury
red pdf green pdf
blue pdf luminosity pdf
2011-09-06
63
Cumulative Distribution Functions (CDF)
red CDF green CDF
blue CDF luminosity CDF
2011-09-06
64
Probability Density Functions of a Color Image
TechnoTrousers
red pdf green pdf
blue pdf luminosity pdf
2011-09-06
65
Cumulative Distribution Functions (CDF)
red CDF green CDF
blue CDF luminosity CDF
2011-09-06
66
Remap an Image to have the Lum. CDF of Another
original
2011-09-06
target
luminosity remapped
67
CDFs and the LUT

Atlas-Mercury Luminosity CDF
TechnoTrousers Luminosity CDF
LUT (Luminosity) Atlas-Mercury to TechnoTrousers
Atlas-Mercury Remapped Luminosity CDF
2011-09-06
68
Effects of Luminance Remapping on pdfs
Before
2011-09-06
After
69
Effects of Luminance Remapping on CDFs
Before
2011-09-06
After
70
Remap an Image to have the rgb CDF of Another
original
2011-09-06
target
R, G, & B remapped
71
CDFs and the LUTs

Atlas-Mercury Red PDF Atlas-Mercury Green PDF Atlas-Mercury Blue PDF
TechnoTrousers Red PDF
TechnoTrousers Green PDF
TechnoTrousers Blue PDF
LUT (Red) Atlas-Mercury to TechnoTrousers
LUT (Green) Atlas-Mercury to TechnoTrousers
LUT (Blue) Atlas-Mercury to TechnoTrousers
Atlas-Mercury RGB Remapped Red PDF
Atlas-Mercury RGB Remapped Green PDF
Atlas-Mercury RGB Remapped Blue PDF
2011-09-06
72
Effects of RGB Remapping on pdfs
Before
2011-09-06
After
73
Effects of RGB Remapping on CDFs
Before
2011-09-06
After
74
Remap an Image:
To Have Two of its Color pdfs Match the Third
original
G&B
B&R
R&G
2011-09-06
75

EECE253 03 PointProcessing

Загружено:

Сведения о документе

Исходное описание:

Оригинальное название

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

EECE253 03 PointProcessing

Загружено:

Авторское право:

Доступные форматы

EECE/CS 253 Image Processing

Point Processing of Images

2000-2011 Richard Alan Peters II

Point Processing of Images

2000-2011 Richard Alan Peters II

The Histogram of a Grayscale Image

2000-2011 Richard Alan Peters II

lower RHC: number of pixels with intensity g

The Histogram of a Grayscale Image

black marks pixels with intensity g

16-level (4-bit) image

2000-2011 Richard Alan Peters II

The Histogram of a Grayscale Image

Plot of histogram: number of pixels with intensity g

2000-2011 Richard Alan Peters II

The Histogram of a Grayscale Image

Plot of histogram: number of pixels with intensity g

2000-2011 Richard Alan Peters II

The Histogram of a Grayscale Image

2000-2011 Richard Alan Peters II

The Histogram of a Color Image

or 1 vector-valued histogram, h(g,1,b) where

2000-2011 Richard Alan Peters II

The Histogram of a Color Image

2000-2011 Richard Alan Peters II

The Histogram of a Color Image

2000-2011 Richard Alan Peters II

Value or Luminance Histograms

Histogram of the value image.

2000-2011 Richard Alan Peters II

Histogram of the luminance image.

2000-2011 Richard Alan Peters II

Value Histogram vs. Average of R,G,&B Histograms

V & avg. of R,G,&B histograms

Multi-Band Histogram Calculator in Matlab

2000-2011 Richard Alan Peters II

Multi-Band Histogram Calculator in Matlab

for g=0:255 h(g+1,1,:) = sum(sum(I==g)); % accumulate end

sum(sum(I==g)) computes one number for each band in the image.

2000-2011 Richard Alan Peters II

Point Ops via Functional Mappings

Point Ops via Functional Mappings

2000-2011 Richard Alan Peters II

Point Ops via Functional Mappings

each band is mapped through a separate function, fb.

Point Operations using Look-up Tables

If k = f ( g ), for g = 0,, 255,

and if k takes on values in {0,, 255},

To remap a 1-band image, I, to J :

2000-2011 Richard Alan Peters II

Point Operations using Look-up Tables

2000-2011 Richard Alan Peters II

Point Operations = Look-up Table Ops

output value 127

127 input value

2000-2011 Richard Alan Peters II

input cell index

a pixel with this value

0 . . . 64 . . . 128 . . . 192 255

0 . . . 32 . . . 128 . . . 224 255

is mapped to this value

2000-2011 Richard Alan Peters II