Spectral Partitioning and its Application to Image Segmentation
Date
2015-07
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of Ghana
Abstract
The properties of graphs can be studied via the algebraic characteristics of its adjacency
or Laplacian matrix. The second eigenvector of the graph Laplacian is one very useful tool
which provides information as to how to partition a graph. In this thesis, we study spectral
clustering and how to apply it in solving the image segmentation problem in computer vision.
Description
Thesis (MPhil)
Keywords
Spectral Partitioning, Application, Image Segmentation