Asymptotic Bayesian Surface Reconstruction

This research is about an iterative algorithm for surface reconstruction based on a sequence of CCD camera images. The iterative algorithm allows the reconstruction process to create a large-baseline disparity field from a sequence of small-baseline disparity fields using the Basian estimatin framework. At each iteration, all the previously estimated information is summarized into a prior term to guide the current estiamtion process. The following image is a snapshot of the the program I developed for this research.

I did this work when I was a master student/research assistant at the Institute of Information Science, Academia Sinica, Taiwan.

Reference: Chun-Jen Tsai, Yi-Ping Hung, and Sung-Ching Hsu, "Comparison between asymptotic Bayesian approach and Kalman filter-based approach for surface reconstruction from a sequence of images," IEEE International Conference on Computer Vision and Pattern Recognition, New York, June, 1993.