Ch02 Fundamentals Matlab

Digital Image Processing Using MATLAB®
Chapter 2
Fundamentals
of
DIP & Matlab
Improve this with www.csie.ntnu.edu.tw/~violet/IP93/Chapter01.ppt

© 2004 R. C. Gonzalez, R. E. Woods, and S. L. Eddins www.imageprocessingbook.com

Chapter 2
Fundamentals

Image Formats
http://en.wikipedia.org/wiki/Image_formats

Displaying Images
>> f = imread(‘rose.tif’);
>> imshow(f)
>> g = imread(‘x-ray.tif’);
>> figure, imshow(g)

>> imshow(f), figure, imshow(g)
>> doc imshow

Displaying Images
>> g = imread(‘xray.tif’);
>> imshow(g)
>> imshow(g, [10 100])
>> imshow(g, [ ])

Displaying Images Info.

>> f = imread(‘rose.tif’);
>> imshow(f)
>> impixelinfo
>> size(f)
ans =
1024 1024
>> [M, N] = size(f);
>> whos f
Name Size Bytes Class Attributes
r 1024x1024 1048576 uint8

Displaying Images Info.

>> imfinfo bubbles.jpg
File size of original image?
ans = Compression ratio?
DPI?
Filename: 'bubble.jpg'
FileModDate: '14-Jan-2008 17:08:08'
Image size?
FileSize: 7904
>> K = imfinfo('bubbles.jpg');
Format: 'jpg'
FormatVersion: ‘'
>> K.FileSize
Width: 720
Height: 688 ans =
BitDepth: 8
ColorType: 'grayscale' 7904
FormatSignature: ‘'
NumberOfSamples: 1
CodingMethod: 'Huffman' Try this
CodingProcess: 'Sequential'
Comment: {} >> imageinfo('rose.tif')
Writing Images
>> b = imread(‘bubbles.tif’);
>> imwrite(b,'bubbles.png')
>> imwrite(b,'bubbles.jpg')
>> imwrite(b,'bubbles', 'jpg')
>> imwrite(b,'bubbles.jpg','quality',50)

Writing Images: * -> tif
'compression': 'none'
'packbits'
'ccitt'
'resolution': dpi
>> ckt = imread('cktboard.jpg');

>> imwrite(ckt,'cktboard-1.tif','compression','none','resolution',[300 300])
>> imwrite(ckt,'cktboard-2.tif','compression','none','resolution',300)
>> imwrite(ckt,'cktboard-3.tif','compression','packbits')
>> imshow(ckt)
>> print
Class Exercise 1
Write Matlab code to get the compression

ratio of an image
Hint: compression ratio = actual size in bites / size on disk in bytes
Actual size in bytes = Height x Width x Bit_depth /8
info = imfinfo(‘imagename.ext’);
Compression_ratio = (info.Height x info.Width x info.BitDepth /8) /
info.FileSize;

Class Exercise 2
What will be the size of the image in

inches if
a) image of size 450 x 450 is saved with 200 dpi
b) image of size 450 x 450 is saved with 300 dpi
a) 2.25 x 2.25 inches

b) 1.5 x 1.5 inches

Data Classes

Converting Between
Data Classes
Syntax: B = data_class_name(A)
Example: Suppose
A is an array of class uint8

C is an array of class double
B = double(A)
D = uint8(C)

Image Types
1. Intensity images
2. Binary images
3. RGB images
4. Indexed images

Converting Between
Image Classes & Types

Converting Between
Example: Consider the following 2 X 2 image f of class double:
>> f = [-0.5 0.5; 0.75 1.5]

f=
-0.5 0.5
0.75 1.5
>> g = im2uint8(f) >> h = uint8([25 50; 128 200]);

>> g = im2double(h);
g= g=
0 128 0.0980 0.1961
191 255 0.4706 0.7843
Converting Between
Example: Conversion of arbitrary array of class double to an array
of class double scaled to the range [0,1]:
g = mat2gray(A, [Amin, Amax])
>> g = mat2gray(A); i.e., Amin = min(A(:)) & Amax = max(A(:))
?
Example:
>> f = [1 2; 3 4]; Can you formulate such a conversion

>> g = mat2gray(f)
g=
0 0.3333
0.6667 1.0000
Converting Between
Conversion between binary and intensity image types:
Syntax:
g = im2bw(f, T) where T has a range [0, 1]
g = im2bw(f, T) T = 0.5 by default
Reading: Example 2.4

Array Indexing
Example:
>> w = v.’
>> v = [1 3 5 7 9] w=
v= 1
1 3 5 7 9 3
>> v(2) 5
ans = 7
3 9
>> v(2:4)
ans =
3 5 7
Guess the output of: v(3:end) v(1:2:end) v(end:-2:1) v([1 4 5])

Matrix Indexing
Example:
>> A = [1 2 3; 4 5 6; 7 8 9] >> A(2:end, end:-2:1)

A= ans =
1 2 3 6 4
4 5 6
7 8 9
9 7
>> A(2, 3) >> C3 = A(:, 3)

ans = C3 =
6 3
6
>> R2 = A(2, :) 9
R2 =
4 5 6
Guess the output of: A(1:2, 1:3) A(end, end-2) A(:, 3) = 0
Matrix Indexing

Matrix Indexing
a.
1024 X 1024 intensity image f of
class uint8
b.
>> fp = f(end:-1:1, :);
c.
>> fc = f(257:768, 257:768);
d.
>> fs = f(1:2:end, 1:2,end);
e.
>> plot(f(512, :))
>> plot(f(size(f, 1)/2, :))

Standard Arrays
>> A = 5*ones(3, 3)
A=
5 5 5
5 5 5
5 5 5
>> magic(3)
ans =
8 1 6
3 5 7
4 9 2
>> B = rand(2, 4)
B=
0.2311 0.4860 0.7621 0.0185
0.6068 0.8913 0.4565 0.8214
M-Function Programming
Components
1. The function definition line: function [outputs] = name(inputs)
2. The H1 line
3. Help text
4. The function body
5. Comments

function [p, pmax, pmin, pn] = improd(f, g) >> help improd
%IMPROD Computes the product of two images.
% [P, PMAX, PMIN, PN] = IMPROD(F, G) >> f = [1 2; 3 4];
>> g = [1 2; 2 1];
fd = double(f); >> [p, pmax, pmin, pn] = improd(f, g)
gd = double(g); p=
p = fd.*gd; 1 4
pmax = max(p(:)); 6 4
pmin = min(p(:)); pmax =
pn = mat2gray(p); % Scaled to [0, 1] 6
pmin =
% The End 1
pn =
0 0.6000
1.0000 0.6000
Exercise:
Write a program to display the dpi and image size

information of any type of image.


Operators
Example: >> A = [1 2 3; 4 5 6; 7 8 9]
A=
1 2 3
f= 4 5 6
1 2 7 8 9
3 4
>> A(2:end, end:-2:1)
ans =
>> v = f(:) 6 4
9 7
v= >> C3 = A(:, 3)
1 C3 =
3 3
6
2 9
4
Arithmetic Operations
Example:
>> imshow(imcomplement(f))
Relational Operators
Example:
>> A = [1 2; 3 4]
A=
1 2
3 4
>> B = [0 2; 3 5;]
B=
0 2
3 5
>> A>=B
ans =
1 1
1 0

Logical Operators and Functions

Functions Returning 1 or 0

Variables & Constants

Flow Control Statements

Code Optimization
1. Vectorizing loops
2. Preallocating arrays
Example: Suppose we want to generate a 1-D
function: >> A=2; M=4;
>> x=0:M-1
f(x) = A sin(x/2) for x = 0, 1, 2,…, M-1 x=
0 1 2 3
Method 1: for x = 1:M >> f = A*sin( x / (2*pi) )
f(x) = A*sin((x-1)/(2*pi)); f=
end 0 0.3170 0.6259 0.9191
Method 2: x = 0:M-1;
f = A*sin(x/(2*pi));

Code Optimization: Example 2.13

function [rt, f, g] = twodsin(A, u0, v0, M, N)
%TWODSIN Compares for loops vs. vectorization.
% The comparison is based on implementing the function
% f(x, y) = Asin(u0x + v0y) for x = 0, 1, 2,...,M-1 and y = 0, 1, 2,...,N-1
% The inputs to the function are M & N and the constants in the function.
% First implement using for loops.
tic % Start timing

for r = 1:M
u0x = u0*(r-1);
for c = 1:N
v0y = v0*(c-1);
f(r,c) = A*sin(u0x + v0y);
end
end
t1 = toc; % End timing.


>> R = 0 : 4
% Now implementing using vectorization. R=
% Call the image g. 0 1 2 3 4
>> C = 0 : 3
tic % Start timing C=
r = 0:M-1; 0 1 2 3
c = 0:N-1;
[C, R] = meshgrid(c, r); >> [C,R] = meshgrid(C,R)
C=
g = A*sin(u0*R + v0*C); 0 1 2 3
t2 = toc; % End timing. 0 1 2 3
0 1 2 3
% Compute the ratio of the two times. 0 1 2 3
0 1 2 3
R=
rt = t1/(t2+eps); % Use eps in case t2 is close to 0. 0 0 0 0
1 1 1 1
2 2 2 2
3 3 3 3
4 4 4 4
>> [rt, f, g] = twodsin(1, 1/(4*pi), 1/(4*pi), 512, 512);

rt =
5.8061 function [rt, f, g] = twodsin(A, u0, v0, M, N)
>> imshow(f)
>> imshow(g)

Interactive I/O
>> t = input('Enter your data: ', 's')
Enter your data: 2 3 4
t=
>> t = input('Enter your data: ', 's')
234
Enter your data: 12.6, x2y, z
>> class(t)
t=
ans =
12.6, x2y, z
char
>> [a, b, c] = strread(t, '%f%q%q', 'delimiter', ',')
n = str2num(t)
a=
>> class(n)
12.6000
ans =
b=
double
'x2y'
>> n(2)
c=
ans =
'z'
3
>> doc strread
>> doc input
String Comparison
Function strcmp is used to compare strings.
Syntax: TF = strcmp('str1', 'str2')
Exercise: Write an M-function g = imnorm(f, param)

that accepts an image, f, and a parameter param (either in
lower or upper case) that can have one of two forms:
‘norm1’: To be scaled to the range [0, 1] with output of

class double
‘norm255’: To be scaled to the range [0, 255] with output
of class uint8

Ch02 Fundamentals Matlab

Hochgeladen von

Dokumentinformationen

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Ch02 Fundamentals Matlab

Hochgeladen von

Copyright:

Verfügbare Formate

Digital Image Processing Using MATLAB®

Improve this with www.csie.ntnu.edu.tw/~violet/IP93/Chapter01.ppt

© 2004 R. C. Gonzalez, R. E. Woods, and S. L. Eddins www.imageprocessingbook.com

© 2004 R. C. Gonzalez, R. E. Woods, and S. L. Eddins www.imageprocessingbook.com

© 2004 R. C. Gonzalez, R. E. Woods, and S. L. Eddins www.imageprocessingbook.com

>> figure, imshow(g)

>> doc imshow

© 2004 R. C. Gonzalez, R. E. Woods, and S. L. Eddins www.imageprocessingbook.com

© 2004 R. C. Gonzalez, R. E. Woods, and S. L. Eddins www.imageprocessingbook.com

Displaying Images Info.

>> [M, N] = size(f);

r 1024x1024 1048576 uint8

© 2004 R. C. Gonzalez, R. E. Woods, and S. L. Eddins www.imageprocessingbook.com

Displaying Images Info.

© 2004 R. C. Gonzalez, R. E. Woods, and S. L. Eddins www.imageprocessingbook.com

Writing Images: * -> tif

>> ckt = imread('cktboard.jpg');

Write Matlab code to get the compression

© 2004 R. C. Gonzalez, R. E. Woods, and S. L. Eddins www.imageprocessingbook.com

What will be the size of the image in

a) 2.25 x 2.25 inches

© 2004 R. C. Gonzalez, R. E. Woods, and S. L. Eddins www.imageprocessingbook.com

© 2004 R. C. Gonzalez, R. E. Woods, and S. L. Eddins www.imageprocessingbook.com

A is an array of class uint8

© 2004 R. C. Gonzalez, R. E. Woods, and S. L. Eddins www.imageprocessingbook.com

© 2004 R. C. Gonzalez, R. E. Woods, and S. L. Eddins www.imageprocessingbook.com

© 2004 R. C. Gonzalez, R. E. Woods, and S. L. Eddins www.imageprocessingbook.com

Example: Consider the following 2 X 2 image f of class double:

>> f = [-0.5 0.5; 0.75 1.5]

>> g = im2uint8(f) >> h = uint8([25 50; 128 200]);

g = mat2gray(A, [Amin, Amax])

>> g = mat2gray(A); i.e., Amin = min(A(:)) & Amax = max(A(:))

>> f = [1 2; 3 4]; Can you formulate such a conversion

Reading: Example 2.4

© 2004 R. C. Gonzalez, R. E. Woods, and S. L. Eddins www.imageprocessingbook.com

Guess the output of: v(3:end) v(1:2:end) v(end:-2:1) v([1 4 5])

>> A = [1 2 3; 4 5 6; 7 8 9] >> A(2:end, end:-2:1)

>> A(2, 3) >> C3 = A(:, 3)

© 2004 R. C. Gonzalez, R. E. Woods, and S. L. Eddins www.imageprocessingbook.com

© 2004 R. C. Gonzalez, R. E. Woods, and S. L. Eddins www.imageprocessingbook.com

© 2004 R. C. Gonzalez, R. E. Woods, and S. L. Eddins www.imageprocessingbook.com

Write a program to display the dpi and image size

© 2004 R. C. Gonzalez, R. E. Woods, and S. L. Eddins www.imageprocessingbook.com

© 2004 R. C. Gonzalez, R. E. Woods, and S. L. Eddins www.imageprocessingbook.com

© 2004 R. C. Gonzalez, R. E. Woods, and S. L. Eddins www.imageprocessingbook.com

Logical Operators and Functions

© 2004 R. C. Gonzalez, R. E. Woods, and S. L. Eddins www.imageprocessingbook.com

© 2004 R. C. Gonzalez, R. E. Woods, and S. L. Eddins www.imageprocessingbook.com

Variables & Constants

© 2004 R. C. Gonzalez, R. E. Woods, and S. L. Eddins www.imageprocessingbook.com

Flow Control Statements

© 2004 R. C. Gonzalez, R. E. Woods, and S. L. Eddins www.imageprocessingbook.com

© 2004 R. C. Gonzalez, R. E. Woods, and S. L. Eddins www.imageprocessingbook.com

Code Optimization: Example 2.13

% First implement using for loops.

tic % Start timing

© 2004 R. C. Gonzalez, R. E. Woods, and S. L. Eddins www.imageprocessingbook.com

Code Optimization: Example 2.13

Code Optimization: Example 2.13

>> [rt, f, g] = twodsin(1, 1/(4*pi), 1/(4*pi), 512, 512);

© 2004 R. C. Gonzalez, R. E. Woods, and S. L. Eddins www.imageprocessingbook.com

Exercise: Write an M-function g = imnorm(f, param)

‘norm1’: To be scaled to the range [0, 1] with output of

Das könnte Ihnen auch gefallen

>> [rt, f, g] = twodsin(1, 1/(4pi), 1/(4pi), 512, 512);