Slides for Mathematically Morphology II - Digital Image Processing | ECE 6258, Study notes of Digital Signal Processing

Material Type: Notes; Class: Digital Image Processing; Subject: Electrical & Computer Engr; University: Georgia Institute of Technology-Main Campus; Term: Fall 2003;

Typology: Study notes

Pre 2010

Uploaded on 08/05/2009

koofers-user-fme-1
koofers-user-fme-1 🇺🇸

10 documents

1 / 7

Toggle sidebar

This page cannot be seen from the preview

Don't miss anything!

bg1
1
11/24/2003 ECE 6258 Russell M. Mersereau 1
ECE6258 Lecture 38
Mathematical Morphology II
Source: Images taken from Gonzalez and Woods.
11/24/2003 ECE 6258 Russell M. Mersereau 2
Extraction of Connected Components
Let Yrepresent a connected component contained
in a set A.
Assume that a point pof Yis known.
All of the elements of Ycan be found by the
following iteration
where X0=p.
Upon convergence Y=Xk.
11/24/2003 ECE 6258 Russell M. Mersereau 3
Connected Component Extraction
11/24/2003 ECE 6258 Russell M. Mersereau 4
Convex Hull
A set Ais convex if a straight line segment drawn between
two points in Alies wholly within A.
The convex hull of Ais the smallest convex set that contains
A.
X = don’t care
pf3
pf4
pf5

Partial preview of the text

Download Slides for Mathematically Morphology II - Digital Image Processing | ECE 6258 and more Study notes Digital Signal Processing in PDF only on Docsity!

ECE 6258 Russell M. Mersereau

ECE6258 Lecture 38

Mathematical Morphology IISource: Images taken from Gonzalez and Woods

ECE 6258 Russell M. Mersereau

Extraction of Connected Components „^

Let

Y

represent a connected component contained

in a set

A

„^

Assume that a point

p

of

Y

is known.

„^

All of the elements of

Y

can be found by the

following iterationwhere

X

p.

„^

Upon convergence

Y

X

. k

ECE 6258 Russell M. Mersereau

Connected Component Extraction

ECE 6258 Russell M. Mersereau

Convex Hull^ „^

A set

A

is

convex

if a straight line segment drawn between

two points in

A

lies wholly within

A

„^

The

convex hull

of

A

is the smallest convex set that contains

A

X = don’t care

ECE 6258 Russell M. Mersereau

Example Note: Convex,but larger thannecessary

ECE 6258 Russell M. Mersereau Correction^ „^ The real convex hull is achieved by limiting thehorizontal and vertical extent to that of the originalobject. 11/24/ ECE 6258 Russell M. Mersereau

Thinning „^ The thinning of a set A by a structuring element B can be defined by „^ This is typically done by thinning with a sequence of structuringelements: {B}={B 1 ,B 2 ,B 3 ,…,B n^ }.

ECE 6258 Russell M. Mersereau Example

ECE 6258 Russell M. Mersereau

Skeleton Example

ECE 6258 Russell M. Mersereau Gray-Scale Dilation „^ Consider a gray-scale function f ( s , t ) and a gray- scale structuring element b ( s , t ). 11/24/ ECE 6258 Russell M. Mersereau

1-D Dilation Example

ECE 6258 Russell M. Mersereau Gray-scale Erosion^ „^ Consider again a gray-scale function f ( s , t ) and a gray-scale structuring element b ( s , t ).

ECE 6258 Russell M. Mersereau

1-D Erosion Example

ECE 6258 Russell M. Mersereau Dilations and Erosions „ Dilations^ ‰ make an image brighter

reduce or eliminate dark details „ Erosions^ ‰ make an image darker

reduce or eliminate bright details „ Dilations and erosions are duals 11/24/ ECE 6258 Russell M. Mersereau

Example (5x5 box) original erosion

dilation

ECE 6258 Russell M. Mersereau Gray-scale Openings and Closings „^ Opening ‰^ Narrow bright areas reduced. „^ Closing ‰^ Narrow dark areas reduced. „^ Duality

ECE 6258 Russell M. Mersereau

Top-Hat Transformation „ „^ Useful for enhancing detail in the presence ofshading

ECE 6258 Russell M. Mersereau Texture Segmentation 11/24/ ECE 6258 Russell M. Mersereau

Texture Segmentation--Algorithm 1.^ Close the input image by using successively largerstructural elements. (This removes the small blobs.)

2.^ A single opening is then performed using a structuringelement that is large compared to the spacing between thelarge blobs. (This leaves a large light region on the left anda large dark region on the right.) 3.^ A simple threshold yields the boundary.

ECE 6258 Russell M. Mersereau Granulometry 1.^ Opening operations with elements of increasing size areperformed.

2.^ Differences between the original and the opening are formed ateach operation. 3.^ A histogram is constructed from the differences. 4.^ The peaks suggest three general particle sizes.