CAREER: Image-Based Rendering using Stereo Vision
Middlebury College, Middlebury VT
Investigators
Abstract
The goal of this research is the application of computer vision techniques, in particular stereo vision techniques, to the domain of image-based rendering (IBR) --- the problem of synthesizing novel views of a scene based on existing images. The long-term goal is to develop a system that builds an image-based scene representation from a stream of input images, while actively selecting new viewpoints. Specific topics of investigation will include: the development of new performance measures for IBR methods; the development of new IBR techniques based on stereo vision; the application of texture synthesis techniques to occlusion issues; the utilization of visual landmarks for calibration; and the development of on-line and off-line algorithms for the selection of reference viewpoints. Undergraduate students will be involved significantly in all aspects of the research project, including implementation, experimentation, authoring papers, and attending conferences. Other students will benefit through the integration of current research topics into the curriculum, and through use of the lab facilities created in the context of the project. The PI's long-term educational goals include developing non-traditional courses aimed at students not ordinarily specializing in the sciences, and establishing contacts to industry for a real-world perspective.
View original record on NSF Award Search →