Segmentation - Part 1

Segmentation – Part 1

Thresholding Segmentation Technique

Elsayed Hemayed

Fall 2007

This material is a modified version of the slides provided by Milan Sonka, Vaclav Hlavac, and Roger
Boyle, Image Processing, Analysis, and Machine Vision..

Computer Vision

Segmentation

Outline



What is Segmentation



Segmentation techniques:

Thresholding

Edge-based Segmentation

Region-based Segmentation

Matching

Computer Vision

Segmentation

Segmentation



Goal:

to divide an image into parts that have a strong

correlation with objects or areas of the real world, mainly
objects and background.



A complete segmentation of an image R is a set of regions
R1, R2, …, Rs



 





R 

i 

Computer Vision

Segmentation

Segmentation



Segmentation may be



Complete segmentation

- set of disjoint regions

uniquely corresponding with objects in the input image



Partial segmentation

- regions do not correspond

directly to image objects

Thresholding

Segmentation Technique



Basic, Band, Multi, Semi-Thresholding



Threshold detection methods



Multi-spectral Thresholding



Hierarchical Thresholding



Algorithms

•

Iterative (optimal) Threshold Selection

•

Recursive multi-spectral thresholding

•

Re-segmenting boundary pixels

Computer Vision

Segmentation

Advantages



simplest

segmentation process since Many

objects or image regions are characterized by
constant reflectivity or light absorption of their
surface.



computationally inexpensive and

fast

and can

easily be done in

real

time

Computer Vision

Segmentation

Basic Thresholding



Basic thresholding of an input image f to an
output image g is as follows:

g(i,j) = 1 for f(i,j)

g(i,j) = 0 for f(i,j)





Matlab Image Processing Toolbox: im2bw
Convert an image to a binary image, based on
threshold

Syntax
BW = im2bw(I,threshold/256);

Computer Vision

Segmentation

Band Thresholding



There are several

modifications

on basic

thresholding. They include:
1.

Band

thresholding:

g(i,j) = 1 for f(i,j)

f(i,j)

g(i,j) = 0 otherwise



used e.g. in microscopic blood cells



Matlab Image Processing Toolbox: roicolor
Select region of interest, based on color

Syntax
BW = roicolor(A,low,high); % 0 <= low, high
<=256

Computer Vision

Segmentation

Multi-Thresholding

2. Multi-thresholding, using limited set of array levels:

g(i,j) = 1

for f(i,j)

g(i,j) = 2

for f(i,j)

g(i,j) = 3

for f(i,j)

.
.
.

g(i,j) = n

for f(i,j)

g(i,j) = 0

otherwise



Computer Vision

Segmentation

Semi-Thresholding

3. Semi-thresholding, masks out background

g(i,j) = f(i,j) for f(i,j)

g(i,j) = 0 for f(i,j)





Computer Vision

Segmentation

Threshold Detection techniques



If some property of an image after segmentation
is known a priori, the task of threshold selection
is simplified, since the threshold is chosen to
ensure this property is satisfied.



A main detection technique is based on

Histogram shape analysis:

•

The chosen threshold is chosen to meet minimum
segmentation error, by selecting it as the gray level
that has

minimum histogram value between the two

maxima

, that represent objects and background in the

image.

Computer Vision

Segmentation

Histogram shape analysis

Distribution of
Background

Distribution of
Objects

Threshold
Value

Matlab Image Processing Toolbox: imhist
Display a histogram of image data

Syntax
imhist(I);

Computer Vision

Segmentation

Histogram shape analysis

(cont.)

Distribution of
Background

Distribution of
Objects

Threshold
Value

Computer Vision

Segmentation

Histogram shape analysis



P-tile thresholding



Weighted histograms



Optimal thresholding

Computer Vision

Segmentation

P-tile thresholding

Threshold is selected such that

1/p

of image pixels

has gray values less than T.

A good example is processing text pages.

1/p

Computer Vision

Segmentation

One option is to

weight histogram

contribution:

•

Excluding

pixels with

high gradient values

that

represent edges will result in histogram with deeper
valley and threshold will be easier to detect.

•

Building histogram for

pixels with high gradient value

only

, the threshold value would be the peak of this

histogram.

Weighted histograms

Computer Vision

Segmentation

Optimal thresholding

It approximates histogram using two or more
probabilities with

normal distribution

. The

threshold is then the min probability between
the maxima of the these normal distributions.

It results in

minimum error segmentation.

Computer Vision

Segmentation

Apple Grading

Results of segmentation by isodata thresholding. Fruits displayed are
defected by scald (top-left), rot (top-right), frost damage (mid-left),
bruise (mid-right), hail damage perfusion (bottom-left) and flesh
damage (bottom-right). For each fruit its original RGB image, its
manual segmentation (ground truth) and its segmentation results are
displayed in a row. Defected areas are displayed in white in ground
truth images, whereas segmentations show defected regions in gray
color and healthy ones in white.

Courtesy of D. UNAY, B. GOSSELIN, 2005,

"Thresholding

-based Segmentation and Apple Grading by Machine Vision"

, Proc. of EUSIPCO 2005, Antalya, Turkey.

Computer Vision

Segmentation

Algorithm:

Iterative (optimal)

Threshold Selection

1. As first iteration consider that the 4 corners contain

background pixels only and the remainder contains

object pixels.

2. Calculate and as the average
intensity of background and object pixels.
3. At step t+1 segmentation is performed using the

threshold



Computer Vision

Segmentation

Iterative (optimal) Threshold

Selection (cont.)

4. Re-calculate and according to new

segmentation using:

5. Re-calculate
6. Stop if



)

(

)

(





An implementation of the Iterative Threshold Selection by
Dhanesh Ramachandram is available online at:

http://www.mathworks.com/matlabcentral/fileexchange/loadFile.do?objectId=
3195&objectType=file

func_threshold

Matlab File

Computer Vision

Segmentation

Initiallize

the whole image as a single region.

Compute a smoothed

histogram for each spectral band

Find the

most significant peak in each

histogram and

determine

two thresholds

on either sides of the peak.

Segment each region

in each spectral band into sub-

regions according to these thresholds.

Algorithm:

Recursive multi-

spectral thresholding

Computer Vision

Segmentation

Project

each segmentation into a multi-spectral

segmentation.

Regions for the next processing

steps are those in the

multi-spectral image.

Repeat

2-6

until

each histogram contains

only one

significat peak.

Recursive multi-spectral

thresholding (cont.):

Computer Vision

Segmentation

The presence of a region is detected in a low
resolution image

Then more precision is given moving to higher
resolution by re-segmenting pixels close to
boundaries.

Advantage:

lower influence of noise!

Algorithm:

Re-segmenting boundary pixels

(

Thresholding in hierarchical data structures)

Computer Vision

Segmentation

Summary



Basic, Band, Multi, Semi-Thresholding



Threshold detection methods



Multi-spectral thresholding



Hierarchical Thresholding



Algorithms

•

Iterative (optimal) Threshold Selection

•

Recursive multi-spectral thresholding

•

Re-segmenting boundary pixels

Document Outline