Two dimensional object recognition using generalized Hough Transform and genetic algorithm

• Main Author: Sim, Hak Chuah.
• Other Authors: Wong, Kok Cheong
• Format: Theses and Dissertations
• Language: English
• Published: 2009
• Subjects: DRNTU::Engineering::Computer science and engineering::Computing methodologies::Image processing and computer vision
• Online Access: http://hdl.handle.net/10356/20508
• Institution: Nanyang Technological University

This thesis focuses on model-based matching which is one of the fundamental components of a general scene interpretation scheme. The model based matching problem is framed within the hypothesis and verification paradigm. Its function is to recognize and to locate instances of a model in a scene based on matching of extracted scene features to stored model features. The obvious and simple approach to obtain the best match is to evaluate all the possibilities, but the incredibly large problem space will take intolerably long computation time to attain a satisfactory result. Current literature reports a number of solutions based on a variety of assumptions and approaches to solve these problems.