Scribd
Upload
34 views

Digital Image Processing-6

I am not sure if you have any questions or need any further information please contact me at the end of the day

Uploaded by

Amir Rehman
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
34 views

Digital Image Processing-6

I am not sure if you have any questions or need any further information please contact me at the end of the day

Uploaded by

Amir Rehman
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 32

Morphological Image Processing (2)

Topics
 Morphological Operations
 Connected Component Extraction
 Convex Hull
 Thinning
 Thickening
 Skeleton
 Pruning
 Extension to gray level images
Dilation and Erosion
 Dilation and Erosion are two basic operations in
morphological processing.
 Dilation of a set A in Z2 by a set B in Z2 is denoted by
A B and given by
Dilation
 The dilation of A by B is the set of all displacements
such that A and overlap with at least one point

 B is called structuring element


Dilation
Erosion
 Erosion of a set A in Z2 by a set B in Z2 is denoted by
and given by:

Erosion of A be B is the set of all points z such that B


translated by z is contained in A
Erosion
Opening
 Opening smoothes the outer contours, breaks narrow
connections, and eliminates small protrusions.

 Opening is defined as :
Closing
 Closing smoothes the object contour, fuses narrow
connections, eliminates small holes and gaps.
Extraction of Connected Components
 Begin with a point P inside the connected component,
iterate:

Until Xk = Xk-1
Initially X0 = P
Connected Component Extraction
Convex Hull
 A set is said convex if the straight line connecting any
two points of the set lies entirely within A.
 Convex Hull of set S is the smallest convex set A that
contains S
 The set difference A-S is called the convex deficiency
of S
Computing Convex Hull
 Let Bi for i=1,2,3,4 represent the structuring elements
shown below
Convex Hull
 Repeat the following equation until converge

with

is the Hit-or-Miss operator

4
Assuming Convex Hull is C ( A)   D i
i 1
Example (Convex Hull)
Improving Convex Hull Algorithm
 The algorithm can be improved by limiting the growth
of the algorithm beyond the maximum dimensions of
the original set.
Thinning and Thickening
 Thinning is an image-processing operation in which
binary valued image regions are reduced to lines
 The purpose of thinning is to reduce the image
components to their essential information for further
analysis and recognition
 Thickening is changing a pixel from 1 to 0 if any
neighbors of the pixel are 1.
 Thickening followed by thinning can be used for
filling undesirable holes.
 Thinning followed by thickening is used for
determining isolated components and clusters.
Thinning
 Thinning is defined in terms of hit or miss as

where B is a sequence of structuring elements like


{B} = {B1, B2, B3, …, Bn} and the operation can be given
as
Thinning
 Sample set of structuring elements
Thinning Example
Thickening
 Thickening is the morphological dual of thinning and
defined as

or
Thickening Example
Skeleton
 The informal definition of a skeleton is a line
representation of an object that is:
 one-pixel thick,
 through the "middle" of the object, and,
 preserves the topology of the object.
Skeleton
 Skeleton is defined by

where

k is the last iterative step before A erodes to an empty


set
Skeleton Example
Pruning
 Thinning and skeletonizing algorithms need a clean-up
post-processing
 The following steps are used for pruning:

 Thinning
 Restore to original form with end points removed

 Dilate end points

 Find the union of X1 and X3


Pruning Example
 Original image and structuring elements
Pruning Example
 Result of thinning and end points detected
Pruning Example
 Dilation of end points and the pruned image
Extension to Gray Level
 Dilation is expressed in 1D as

 Erosion is given by
Extension to Gray Level (2D Case)
 Dilation

 Erosion
Questions?

You might also like