Computer Vision Talks at Columbia University

Real-Time Motion Analysis with Linear-Programming

Moshe Ben Ezra

Columbia University

CAVE Lab, 6th Floor CEPSR 

Host: Prof. Shree Nayar 

Abstract 

This is Joint Work with Prof. Shumel Peleg and Prof. Michael Werman Hebrew University of Jerusalem.

In this talk, a method to compute motion models in real time from point-to-line correspondences using linear programming is presented. Point-to-line correspondences are used since they are reliable measurements for image motion given the aperture effect, and it is shown how they can approximate other motion measurements as well.

An error measure for image alignment using the L1 metric, and based on point-to-line correspondences, achieves results which are more robust than the commonly used L2 metric. The L1 error measure is minimized using linear programming. While estimators based on L1 are not robust in the breakdown point sense, experiments show that the proposed method is robust enough to allow accurate motion recovery over long video sequences in real-time.