This is the first year funding of a four-year continuing award. The research component of this CAREER program will focus on the problem of constructing an object level, 3D model of a scene using the image data obtained from one or more video cameras mounted on a mobile robot platform. The ability to construct and maintain accurate, object level models of the environment registered to the image data is a fundamental capability that could be exploited in a variety of ways in a mobile robot application. Toward this end, research will be carried out on three related problems: (1) extending the reconstruction techniques to make use of the imagery acquired with an omnidirectional camera system (2) developing effective model tracking algorithms that can be used to update the model of the scene as the robot moves (3) developing algorithms that can automatically determine how a given scene could be modeled in terms of parameterized primitives. The education component of this award includes new course curriculum development, undergraduate and graduate research projects, mentoring of minority students and outreach in the K-12 system. Outreach efforts to minority high-school students will include the development of new course curriculum modules which take advantage of students' natural curiosity about robotic systems to reinforce learning of fundamental techniques in mathematics, physics and computer science.
