A novel vision system for baggage localization

Abstract

Given the challenges of manual operation, there is an impending need for automation of the baggage handling problem. Imparting sensory capabilities is important in the creation of intelligent robotic systems that can achieve this task. In this paper, we focus on vision as it is one of the dominant perceptual modalities. Specifically, we propose a part based edge alignment technique based on Chamfer Matching for estimating the 2D localization of baggage in an uncluttered scene. This novel vision system incorporates the power and flexibility of both object-centered and viewer-centered theories of object representation.