Fundamentals of Image Processing & Computer Vision with MATLAB

By
Ali Ghanbarzadeh
ghanbarzade.ali@gmail.com
1

 Introduction to MATLAB
 Image Processing
 Image Processing MATLAB Toolbox
 Image Processing Examples
 Computer Vision
 Computer Vision MATLAB Toolbox
 Computer Vision Examples
 Applications
 Alternatives to MATLAB
Overview
2

MATLAB (matrix laboratory) is a multi-paradigm numerical computing
environment and fourth-generation programming language developed by
MathWorks.
MATLAB allows matrixmanipulations, plotting of functions and data,
implementation of algorithms, creation of user interfaces, and interfacing with
programs written in other languages, including C, C++, Java, Fortran and Python.
Introduction to MATLAB
3

MATLAB User Interface
• Command Window
• Current Folder
• Workspace
4

>> clc
Clears command window
>> help
Displays help text in Command Window
>> clear
Clears variables and functions from memory.
removes all variables from the workspace.
>> clear ‘a variable/variables’
5

>> ver
Displays MathWorks product family header information, followed by the current MATLAB,
Simulink and toolbox version information
to write comments use %
>> %This is a comment
6

Basic Arithmatics ( + , - , * , / )
Variables
every variable is an array or matrix
>> x=10
Creates a 1-by-1 matrix named x and stores the value 10 in its element.
Once a variable is entered into the system, you can refer to it later.
Variables must have values before they are used.
7

When an expression returns a result that is not assigned to any variable, the system
assigns it to a variable named ans, which can be used later.
>> 212/23
ans =
9.2174
You can use this variable ans
>> ans+5
ans =
14.2174
8

 who command displays all the variable names you have used
>> who
Your variables are:
a ans b x
whos command gives more information
>> whos
Name Size Bytes Class
a 1x1 8 double
ans 1x1 8 double
b 1x1 8 double
x 1x1 8 double
9

Constants
>> pi
ans =
3.1416
>> eps
ans =
2.2204e-16
 Symbolic Math Toolbox
>> x=1/3
x =
0.3333
>> x=sym(1/3)
x =
1/3
>> double(x)
ans =
0.3333 10

Trigonometric Functions
>> sin(pi/6)
ans =
0.5000
>> cos(0) + tan(pi/4)
ans =
2
11

>> exp(1)
ans =
2.7183
>> log(exp(1))
ans =
1
>> log10(1000)
ans =
3
>> 4^3
ans =
64
>> sqrt(81)
ans =
9
>> nthroot(625,4)
ans =
5
>> log2(1024)
ans =
10
>> factorial (5)
ans =
120
12

>> m(3)
ans =
3
>> length(m)
ans =
5
>> sum(m)
ans =
15
>> m=[1 2 3 4 5]
m =
1 2 3 4 5
>> m'
ans =
1
2
3
4
5
>> min(m)
ans =
1
>> max(m)
ans =
5
>> mean(m)
ans =
3
 Vectors
13

>> a.*b
ans =
3 16 35 60 91
>> p=a.^2
p =
1 16 49 100 169
>> sqrt(p)
ans =
1 4 7 10 13
 Vectors
>> a=1:3:15
a =
1 4 7 10 13
>> a(2:5)
ans =
4 7 10 13
>> b=(3:7)
b =
3 4 5 6 7
14

>> n(2,3)
ans =
4
>> n(3,:)
ans =
7 8 6
>> max(n)
ans =
7 9 6
>> min(n)
ans =
2 5 1
>> sum(n)
ans =
12 22 11
>> sum(sum(n))
ans =
45
 Matrices
>> n=[3 5 1;2 9 4;7 8 6]
n =
3 5 1
2 9 4
7 8 6
>> n(1:3,2:3)
ans =
5 1
9 4
8 6
15

>> diag(n)
ans =
3
9
6
>> det(n)
ans =
99.0000
>> triu(n)
ans =
3 5 1
0 9 4
0 0 6
>> tril(n)
ans =
3 0 0
2 9 0
7 8 6
 Matrices
>> numel(n)
ans =
9
>> size(n)
ans =
3 3
>> n'
ans =
3 2 7
5 9 8
1 4 6
16

>> (30-20)*(rand(5,1))+20
ans =
21.5761
29.7059
29.5717
24.8538
28.0028
>> ones(2)
ans =
1 1
1 1
>> zeros(3)
ans =
0 0 0
0 0 0
0 0 0
>> eye(3)
ans =
1 0 0
0 1 0
0 0 1
 Matrices
>> rand(2)
ans =
0.8147 0.1270
0.9058 0.9134
>> rand(3,2)
ans =
0.6324 0.5469
0.0975 0.9575
0.2785 0.9649
17

>> std(data)
ans =
26.0109
>> var(data)
ans =
676.5667
 Statistics
>> data=[12,54,23,69,31,76]
data =
12 54 23 69 31 76
>> sort(data)
ans =
12 23 31 54 69 76
>> median(ans)
ans =
42.5000
18

Functions
>> area_rectangle=inline('2*(a+b)','a','b')
area_rectangle =
Inline function:
area_rectangle(a,b) = 2*(a+b)
>> area_rectangle(5,7)
ans =
24
function max = mymax(n1, n2, n3)
max = n1;
if(n2 > max)
max = n2;
end
if(n3 > max)
max = n3;
end
disp(max);
19

Calculus – differentiation
>> syms x
>> f=inline('sin(x)+2*cos(x)','x')
f =
Inline function:
f(x) = sin(x)+2*cos(x)
>> diff(f(x),x)
ans =
cos(x) – 2*sin(x)
>> syms x y
>> f=inline('sin(x)+cos(y)','x','y')
f =
Inline function:
f(x,y) = sin(x)+cos(y)
>> diff(f(x,y),x)
ans =
cos(x)
20

Calculus - integral
>> syms x
>> f=inline('x/sin(x^2)','x')
f =
Inline function:
f(x) = x/sin(x^2)
>> integral=int(f(x),x)
integral =
- log(- x*2*i - x*exp(x^2*i)*2*i)/2 + log(x*2*i - x*exp(x^2*i)*2*i)/2
>> pretty(integral)
2 2
log(- x 2 i - x exp(x i) 2 i) log(x 2 i - x exp(x i) 2 i)
- ------------------------------ + ----------------------------
2 2
21

Calculus – integral
>> syms x
>> f=inline('(x^3)+2','x')
f =
Inline function:
f(x) = (x^3)+2
>> integral=int(f(x),x,2,4)
integral =
64
22

Calculus – limit
>> syms x
>> f=inline('sin(x)/x','x')
f =
Inline function:
f(x) = sin(x)/x
>> limit(f(x),x,0)
ans =
1
>> syms x
>> g=inline('(x+x^2)/((2*x^2)+3*x) ','x')
g =
Inline function:
g(x) = (x+x^2)/((2*x^2)+3*x)
>> limit(g(x),x,Inf)
ans =
1/2
23

Plotting
>> x=[4 6 8 11 12];
>> y=[1 3 3 5 7];
>> plot(x,y)
>> bar(y)
>> pie(x)
24

input
>> a=input('enter value for array: ')
enter value for array: [3 4 9 0 2]
a =
3 4 9 0 2
>> name=input('enter your name: ','s')
enter your name: Ali
name =
Ali
>> whos
Name Size Bytes Class
name 1x19 38 char
25

Decisions
>> a = 100;
if a < 20
disp('less than 20');
else
disp('not less than 20');
end
disp(a);
not less than 20
100
>> grade = 'B';
switch(grade)
case 'A'
disp('A');
case 'B'
disp('B');
case 'C'
disp('C');
otherwise
disp('Invalid grade');
end
26

Loops
>> for a = 1.0: -0.1: 0.0
disp(a)
end
1
0.9000
0.8000
….
0.1000
0
>> a = 10;
while( a < 20 )
disp(a)
a = a + 2;
end
10
12
14
16
18
27

Image Processing is processing of images using mathematical operations by using
any form of signal processing for which the input is an image, a series of images, or a
video, and the output of image processing may be either an image or a set of
characteristics or parameters related to the image.
Image processing usually refers to digital image processing, but optical and analog
image processing also are possible.
Closely related to image processing are computer graphics and computer vision. In
computer graphics, images are manually made from physical models of objects
instead of being acquired via imaging devices. Computer vision is considered
high-level image processing out of which a machine/computer/software intends to
decipher the physical contents of an image or a sequence of images.
Image Processing
28

Features
Image Processing Toolbox supports a diverse set of image types
image analysis
image segmentation
image enhancement
 noise reduction
geometric transformations
Visualization functions and apps let you
explore images and videos,
examine a region of pixels, adjust color and contrast,
create contours or histograms
Image Processing MATLAB Toolbox
29

The basic data structure in MATLAB is the array.
MATLAB stores most images as two-dimensional arrays (i.e., matrices), in which
each element of the matrix corresponds to a single pixel in the displayed image.
Images in MATLAB
30

Some images, such as truecolor images,
require a three-dimensional array, where
the first plane in the third dimension
represents the red pixel intensities,
the second plane represents the green
pixel intensities, and the third plane
represents the blue pixel intensities.
Images in MATLAB
31

I=imread('balls.jpg');
imtool(I);
Exploring Images (Measurement)
32

Exploring Images (Inspect Pixels Value)
33

An image histogram is a type of histogram that acts as a graphical
representation of the tonal distribution in a digital image.
The horizontal axis of the graph represents the tonal variations, while the
vertical axis represents the number of pixels in that particular tone.
Histogram of Image
34

Clc;clear;
I = imread('pout.tif'); %reading an image
figure;
subplot(2,2,1);
imshow(I); %shows the image
subplot(2,2,2);
imhist(I); %Views the distribution of image pixel intensities
I2 = histeq(I); %improves the contrast in an image,using the histeq
%function.
subplot(2,2,3);
imshow(I2);
subplot(2,2,4);
imhist(I2);
Enhancement (1)
35

imwrite (I2,'pout2.png','png'); %Writes the newly adjusted image to a
%file.
imfinfo('pout2.png'); %Gives the information about the image
Enhancement (1)
36

clc; clear;
I = imread('pout.tif'); %reading an image
figure;
subplot(2,2,1)
imshow(I) %shows the image
subplot(2,2,2)
imhist(I) %Views the distribution of image pixel intensities
I3=imadjust(I); %Stretches the histogram
subplot(2,2,3)
imshow(I3)
subplot(2,2,4)
imhist(I3)
Enhancement (2)
37

histeq
By default, MATLAB transforms the probability distribution to a uniform
distribution, which means all the intensity ranges have equivalent probabilities.
The histogram is modified to have the counts of all pixels close to each other
(uniform distribution)
imadjust
imadjust just "Stretches" the histogram of an image in order to have the
intensity range of the image fill the entire available range (normally 0-255).
histeq vs. imadjust
39

clc; clear;
I=imread('cameraman.tif');
subplot(1,3,1)
imshow(I)
title('original')
V= I(end:-1:1,:); % Vertical Flip
subplot(1,3,2)
imshow(V)
title('vertical')
Flip
40

H= I(:,end:-1:1); % Horizontal Flip
subplot(1,3,3)
imshow(H)
title('Horizontal')
Flip
41

scale = 0.5;
R = imresize(I,scale); % Resizing
figure; imshow(R)
title('Resized')
theta = 35;
Rotation = imrotate(I,theta); % Rotating
figure; imshow(Rotation)
title('Rotated')
Imwrite(Rotation,'r.png','png') % writes image data to the file
% specified by filename
Rotate, Resize
42

Thresholding is the simplest method of image segmentation.
From a grayscale image, thresholding can be used to create binary
images.
The simplest thresholding methods replace each pixel in an image
with a black pixel if the image intensity 𝐼𝑖,𝑗 is less than some fixed
constant T(that is 𝐼𝑖,𝑗 < T), or a white pixel if the image intensity is
greater than that constant.
Thresholding
44

The approach for thresholding color images is to designate a separate
threshold for each of the RGB components of the image and then combine
them with an AND operation.
Thresholding
45

clc; clear;
I=imread('balls.jpg');
red=I(:,:,1); %extracts the red channel
green=I(:,:,2); %extracts the green channel
blue=I(:,:,3); %extracts the blue channel
figure;
subplot(1,3,1); imshow(red); title('red');
subplot(1,3,2); imshow(green); title('green');
subplot(1,3,3); imshow(blue); title('blue');
Thresholding
46

figure;
pix=impixel(I); %the tool for getting pixel information
Thresholding
48

maxPixels=max(pix); %finds the maximum of columns
minPixels=min(pix); %finds the minimum of columns
out = red<=maxPixels(1) & red>=minPixels(1) &...
green<=maxPixels(2) & green>=minPixels(2) &...
blue<=maxPixels(3) & blue>= minPixels(3);
%trying to thershold the selected segment
figure;
imshow(out);
Thresholding
49

out2 = imfill(out,'holes'); % filling the holes
figure;
imshow(out2);
title('without holes');
stats = regionprops(out2); %information about the segment
Thresholding
51

In signal processing, a filter is a device or process that removes from a
signal some unwanted component or feature.
fspecial
Creates predefined 2-D filter
Filters
53

medfilt2
2-D median filtering
The median filter is a nonlinear digital filtering technique, often used to
remove noise.
Filters
54

clc; clear;
I=imread('cat.png');
figure;
imshow(I);title('Original Image');
motion_f=fspecial('motion',10);
I2=imfilter(I,motion_f);
figure;
imshow(I2);title('Blurred Image');
I3=imsharpen(I2,'amount',5);
figure;imshow(I3);title('Sharpened Image');
Sharpening
55

clc; clear;
I=imread('mo.jpg');
I=rgb2gray(I);
figure;imshow(I);
Filter Noise
57

N=imnoise(I,'salt & pepper',0.25);
figure,imshow(N);
average_f=fspecial('average');
Filtered=imfilter(N,average_f);
figure,imshow(Filtered);
Filter Noise
58

Filtered=medfilt2(N);
figure,imshow(Filtered);
Filter Noise
59

clc; clear;
a=imread('shelf.png');
b=rgb2gray(a);
figure; imshow(a);
h=[-1 -1 -1;2 2 2; -1 -1 -1]; %filter for horizontal lines
c1=imfilter(b,h);
figure; imshow(c1);
h=[-1 2 -1;-1 2 -1; -1 2 -1]; %filter for vertical lines
c2=imfilter(b,h);
figure; imshow(c2);
Line Detection
60

d=imadd(c1,c2); %adding two images
figure; imshow(d);
Line Detection
62

clc; clear;
a=imread('house.jpg');
b=rgb2gray(a); %converts the RGB image to grayscale
figure; imshow(b);
Edge Detection
63

e=edge(b,'sobel'); %edge detection with 'sobel'
figure; imshow(e);
Edge Detection
64

e=edge(b,'canny'); %edge detection with 'canny'
figure; imshow(e);
Edge Detection
65

h=fspecial('laplacian');
e=imfilter(b,h);
figure; imshow(e);
Edge Detection
66

Computer vision is a field that includes methods for acquiring, processing,
analyzing, and understanding images and, in general, high-dimensional data
from the real world in order to produce numerical or symbolic information,
e.g., in the forms of decisions.
As a scientific discipline, computer vision is concerned with the theory
behind artificial systems that extract information from images. The image
data can take many forms, such as video sequences, views from multiple
cameras, or multi-dimensional data from a medical scanner.
Computer Vision
67

Computer Vision System Toolbox provides algorithms, functions, and apps for the
design and simulation of computer vision and video processing systems. You can
perform object detection and tracking, feature detection and extraction, feature
matching, camera calibration, and motion detection tasks.
The system toolbox also provides tools for video processing, including video file I/O,
video display, object annotation, drawing graphics.
Computer Vision MATLAB Toolbox
68

clc; clear;
A = imread('circlesBrightDark.png');
figure;imshow(A);
Rmin = 30;
Rmax = 65;
[centersBright, radiiBright] = imfindcircles(A,[Rmin …
Rmax],'ObjectPolarity','bright');
[centersDark, radiiDark] = imfindcircles(A,[Rmin …
Rmax],'ObjectPolarity','dark');
figure;imshow(A);
viscircles(centersBright, radiiBright,'EdgeColor','b');
viscircles(centersDark, radiiDark,'LineStyle','--');
Find Bright & Dark Circles
69

Optical character recognition (optical character reader) (OCR) is the mechanical
or electronic conversion of images of typed, handwritten or printed text into
machine-encoded text.
data entry from printed paper data records, whether passport documents,
invoices, bank statements, computerised receipts, business cards, mail, printouts
of static-data, or any suitable documentation.
used in machine processes such as machine translation, text-to-speech and text
mining.
OCR is a field of research in pattern recognition, artificial intelligence and
computer vision.
OCR (Optical Character Recognition)
71

clc; clear;
I = imread('ocr_test.png');
ocrResults = ocr(I); %OCR for getting the text out of an image
recognizedText = ocrResults.Text; %the Text field
figure;
imshow(I);
disp(recognizedText) %displays the whole text in Command Window
OCR
72

boxed = insertObjectAnnotation(I,'rectangle',…
ocrResults.WordBoundingBoxes,…
ocrResults.WordConfidences);
%showing the boxes around the words with the
%relevant word confidence
figure;
imshow(boxed)
OCR
74

OCR
Ran Zhang
Xuemin (Sherman) Shen
Mobile Electric
Vehicles
Online Charging and
Discharging
75

The cascade object detector uses the Viola-Jones algorithm to detect people's faces,
noses, eyes, mouth, or upper body.
detector = vision.CascadeObjectDetector creates a System object, detector, that
detects objects using the Viola-Jones algorithm. The ClassificationModel property
controls the type of object to detect. By default, the detector is configured to detect
faces.
BBOX = step(detector,I) returns BBOX, an M-by-4 matrix defining M bounding boxes
containing the detected objects. This method performs multiscale object detection on
the input image, I. Each row of the output matrix, BBOX, contains a four-element
vector, [x y width height], that specifies in pixels, the upper-left corner and size of a
bounding box. The input image I, must be a grayscale or truecolor (RGB) image.
Cascade Object Detection
76

clc; clear;
for c=1:7
I=imread(num2str(c),'jpg');
detector=vision.CascadeObjectDetector;
box = step(detector, I);
out = insertObjectAnnotation(I,'rectangle',box,'face');
figure; imshow(out);
end
Face Detection
77

clc; clear;
I1=imread('b2.png');
I2=imread('sm.jpg');
detector=vision.CascadeObjectDetector('EyePairBig');
detector.MergeThreshold=10;
box1 = step(detector, I1);
out1 = insertObjectAnnotation(I1,'rectangle',box1,'eyes');
figure; imshow(out1);
out2 = insertObjectAnnotation(I2,'rectangle',box2,'eyes');
Face Feature Detection(Eyes)
79

Face Feature Detection(Eyes)
80

clc; clear;
detector=vision.CascadeObjectDetector('Mouth');
out1 = insertObjectAnnotation(I1,'rectangle',box1,'mouth');
out2 = insertObjectAnnotation(I2,'rectangle',box2,'mouth');
Face Feature Detection(Mouth)
81

Face Feature Detection(Mouth)
82

clc; clear;
detector=vision.CascadeObjectDetector('Nose');
out1 = insertObjectAnnotation(I1,'rectangle',box1,'nose');
out2 = insertObjectAnnotation(I2,'rectangle',box2,'nose');
Face Feature Detection(Nose)
83

Face Feature Detection(Nose)
84

clc; clear;
info = imaqhwinfo('winvideo');
info.DeviceInfo(1)
info.DeviceInfo.SupportedFormats
vid = videoinput('winvideo',1,'YUY2_640x480');
preview(vid)
Stop(vid)
Video Acquisition
85

clc; clear;
faceDetector = vision.CascadeObjectDetector();
obj =imaq.VideoDevice('winvideo', 1, 'YUY2_640x480','ROI', …
[1 1 640 480]);
set(obj,'ReturnedColorSpace', 'rgb');
figure('menubar','none','tag','webcam');
Face Detection from live video stream
86

while (true)
frame=step(obj);
bbox=step(faceDetector,frame);
boxInserter =
vision.ShapeInserter('BorderColor','Custom',...
'CustomBorderColor',[255 255 0]);
videoOut = step(boxInserter, frame,bbox);
imshow(videoOut,'border','tight');
f=findobj('tag','webcam');
87

if (isempty(f));
[hueChannel,~,~] = rgb2hsv(frame);
hold off
noseDetector = vision.CascadeObjectDetector('Nose');
faceImage = imcrop(frame,bbox);
noseBBox = step(noseDetector,faceImage);
videoInfo = info(obj);
ROI=get(obj,'ROI');
VideoSize = [ROI(3) ROI(4)];
tracker = vision.HistogramBasedTracker;
initializeObject(tracker, hueChannel, bbox);
release(obj);
close(gcf)
break
end % End of if
pause(0.05)
end % End of While
88

Pattern Recognition is a branch of machine learning
that focuses on the recognition of patterns and
regularities in data, although it is in some cases
considered to be nearly synonymous
with machine learning.
Pattern Recognition algorithms generally aim
to provide a reasonable answer for all possible
inputs and to perform "most likely" matching of
the inputs, taking into account their statistical variation.
Pattern Recognition
89

Biometric Image Processing and Recognition
Applications
90

Biometric Image Processing and Recognition
Applications
91

Medical Imaging
Applications
92

Surveillance
Applications
93

Industrial Control
Applications
94

Computer Aided Surgery
Applications
96

Geospatial Computing
Applications
97

Mathematica
GNU Octave
R (programming language)
Scilab
SciPy & NumPy
Alternatives to MATLAB
99

Fundamentals of Image Processing & Computer Vision with MATLAB

More Related Content

What's hot

Similar to Fundamentals of Image Processing & Computer Vision with MATLAB

Recently uploaded

Fundamentals of Image Processing & Computer Vision with MATLAB