Introduction to Image Processing in Matlab

Introduction to Image Processing in Matlab & Binary Image Analysis

Welcome to your first Computer Vision Lab.  This lab will introduce you to writing

image processing software using Matlab and the Matlab Image Processing Toolbox.  You will learn how to load, view and manipulate images, convert between image formats,  threshold images, and you will write a function that calculates the moments and orientation of a binary object.  

Deliverable:

To gain the marks for this lab you will need to show me your Lab1 function running

during the lab.  The specification for this function is towards the end of this document.  

To get started quickly work through the Getting Started and Introduction to Image

Processing in Matlab sections.

Getting Started

1.   Make a new directory in which to do your work for this course.

2.   Download the image ‘text.tiff’ from

http://www.syseng.anu.edu.au/~luke/cvcourse_files/images/text.tiff 

Save this image in your new directory.

3.   Open Matlab from the Start>Programs menu.

4.   Change to your new directory (use cd).

5.   Type edit to open a new script file, and you are ready to start work. 

I suggest you do all your work in your script file. 

Did you know:

 •  You can write multiple Matlab functions in the same file?  This is helpful to keep your work together especially when working on large projects, to do this you will have to make the file a function. 

•  You can use the debugging features in the Matlab editor to jump between the local variable spaces when debugging your functions.  Check 'Stop if Error' in the Breakpoints menu of the Matlab editor to activate the debugger on the next error.  Note the different cursor in debugging mode.  

•  dbquit exits debugging mode.  Remember to quit debugging mode before

rerunning your code.

Introduction to Image Processing in Matlab

You can get help on any Matlab function by typing help <function> or to get help

specifically on the Image P rocessing Toolbox type help images, typing just help

will show you all the available help topics.

Load the image into im_rgb with

im_rgb = imread(‘text.tiff’);

convert to type double

im_rgb = double(im_rgb);

convert to grey-scale

im_grey = (im_rgb(:,:,1)+im_rgb(:,:,2)+im_rgb(:,:,3))/3;

or you can use rgb2gray which gives a slightly different result since it converts to a

luminance rather than an  intensity image, you'll learn about this in the lecture on colour imaging.  For this lab it doesn't matter which one you use.

convert from [0,255] to [0,1] 

im_grey = im_grey/255;

to view an image use either imshow(im) or imagesc(im), note the difference.  You

can use imagesc and then set the axis for an image with 

axis image ;

Axis labelling can be turned  on and off by  

axis on; 

axis off;

You can interactively crop an image using imcrop(im_grey); Try this.  For this lab

exercise I would like you to crop the image to a specified region containing only the 2nd letter.  Do this with

im_grey = im_grey(150:270, 280:400);

Examine the intensity profile of im_grey

improfile(im_grey);

Threshold im_grey

im_bin = imgrey>0.5; 

The object we are interest in is the text, since we have defined binary objects to be 1’s  and the background to be zero we need to invert the image

im_bin = 1 – im_bin;   % inverts the binary image.

You can find the coordinates of the points in this object using  find.  Type help find

to learn about this function.  Write a line of code using  find that returns the x and y

values of all the points in the object.