Committee: 

Dr. AlRegib, Advisor    

Dr. Anderson, Chair

Dr. Jayaraman

Abstract:

The objective of the proposed research is designing a texture analysis framework and algorithms for material characterization, motion surveillance, and quality inspection in advanced manufacturing vision systems using handcrafted and machine-learning- texture representation methods. Our framework performs three main vision tasks: (1) extracting representative texture features and distinguishing various material types, (2) recognizing periodically repeat texture patterns exhibited in fabrics, and (3) tracking the position of materials for motion surveillance. For stochastic texture without patterns, we develop algorithms to examine material characteristics.