Journal of Theoretical and Applied Information Technology
© 2005 - 2010JATIT. All rights reserved.
www.jatit.org
1
NOVEL GRAPH BASED METHOD FOR IMAGE
SEGMENTATION
1
Dr. S. V.KASMIR RAJA,
2
A. SHAIK ABDUL KHADIR,
3
Dr. S. S. RIAZ AHAMED
1
Dean (Research), SRM University, Chennai, TamilNadu, India
2
Lecturer (SG), Dept of Computer Science, Khadir Mohideen College, Adirampattinam-614701,
TamilNadu, India-614701
3
Principal, Sathak Institute of Technology, Ramanathapuram,TamilNadu, India-623501.
Email: ssriaz@ieee.org , ssriaz@yahoo.com
ABSTRACT
We propose a novel approach for solving the perceptual grouping problem in vision. Rather than focusing
on local features and their consistencies in the image data, our approach aims at extracting the global
impression of an image. We treat image segmentation as a graph partitioning problem and propose a novel
global criterion, the normalized cut, for segmenting the graph. The normalized cut criterion measures both
the total dissimilarity between the different groups as well as the total similarity within the groups .We
show that an efficient computational technique based on a generalized eigen value problem can be used to
optimize this criterion. At the heart of unsupervised clustering and semi-supervised clustering is the
calculation of matrix Eigen values (eigenvectors) or matrix inversion. In generally, its complexity is O(N
3
).
By using Fast Lanczos Method in Normalized cut Method, we improve the performance to O(N log N).
We have applied this approach to segmenting static images, as well as motion sequences, and found the
results to be very encouraging.
Keywords : Grouping, image segmentation, graph partitioning., unsupervised clustering
1. INTRODUCTION
Nearly 75 years ago, Wertheimer [1] pointed
out the importance of perceptual grouping and
organization in vision and listed several key
factors, such as similarity, proximity, and good
continuation, which lead to visual grouping.
However, even to this day, many of the
computational issues of perceptual grouping
have remained unresolved. In this paper, we
present a general framework for this problem,
focusing specifically on the case of image
segmentation.
Prior literature on the related problems of
clustering, grouping and image segmentation is
huge. The clustering community [3] has offered
us agglomerative and divisive algorithms; in
image segmentation, we have region-based
merge and split algorithms. The hierarchical
divisive approach that we advocate produces a
tree, the dendrogram. While most of these ideas
go back to the 1970s (and earlier), the 1980s
brought in the use of Markov Random Fields [2]
and variational formulations [6], [4], [5].
Data Clustering and graph segmentation
are important operations for machine learning
and computer vision. It includes both
unsupervised and semi-supervised clustering.
For unsupervised clustering, spectral
method[8][9][10][7] has been the focus of
considerable research. However, all these
methods suffer from the slow
computation of large matrix. Solving
systems of linear equations Ax = b and
computing eigenvalues and eigenvectors of large
matrices Ax=λx are two fundamental
computation for clustering problem. They also
have great practical uses in other machine
learning problems. For example, for semi-
supervised clustering, to label unlabeled points
can be transformed into a problem of solving
linear systems. And for the spectral clustering
problem, it turns out to be a problem of
computing the largest k eigenvectors of the
weighted matrix W. Hence how to solve these
two problems correctly and fast becomes very
important.