Date of Submission
Summer 7-29-2016
Degree Type
Thesis
Degree Name
Master of Science in Computer Science (MSCS)
Department
Computer Science
Committee Chair/First Advisor
Chih-Cheng Hung
Track
Others
Computer Vision
Chair
Frank Tsui
Committee Member
Jeffrey Chastine
Committee Member
Chia-Tien Dan Lo
Abstract
This thesis presents color image segmentation as a vital step of image analysis in computer vision. A survey of the Markov Random Field (MRF) with four different implementation methods for its parameter estimation is provided. In addition, a survey of swarm intelligence and a number of swarm based algorithms are presented. The MRF model is used for color image segmentation in the framework. This thesis introduces a new image segmentation implementation that uses the bee algorithm as an optimization tool in the Markovian framework. The experiments show that the new proposed method performs faster than the existing implementation methods with about the same segmentation accuracy.