Content based video indexing and retrieval using motion features

Abstract

Presented in this thesis is MPEG VIRS: A content based video indexing and retrieval system. It is a prototype system that allows the automatic population and indexing of a database of MPEG-1 video files, based on colour, texture, and motion features. The system also allows searching of the database via a graphical sketch. The main focus of this thesis is the implementation of the motion indexing process, and additionally the integration of this component into the system as a whole. MPEG VIRS is designed to display the potential of a content based indexing and retrieval system in the growing need to provide tools to aid the management of large video databases, as well as to provide a framework for further research into this area.