Title :
Fast, robust, and consistent camera motion estimation
Author :
Zhang, Tong ; Tomasi, Carlo
Author_Institution :
Dept. of Comput. Sci., Stanford Univ., CA, USA
Abstract :
Previous algorithms that recover camera motion from image velocities suffer from both bias and excessive variance in the results. We propose a robust estimator of camera motion that is statistically consistent when image noise is isotropic. Consistency means that the estimated motion converges in probability, to the true value as the number of image points increases. An algorithm based on reweighted Gauss-Newton iterations handles 100 velocity measurements in about 50 milliseconds on a workstation
Keywords :
motion estimation; noise; probability; velocity measurement; camera motion; consistent camera motion estimation; image noise; image points; image velocities; reweighted Gauss-Newton iterations; robust estimator; velocity measurements; Cameras; Image converters; Least squares methods; Motion estimation; Newton method; Noise robustness; Probability; Recursive estimation; Velocity measurement; Workstations;
Conference_Titel :
Computer Vision and Pattern Recognition, 1999. IEEE Computer Society Conference on.
Conference_Location :
Fort Collins, CO
Print_ISBN :
0-7695-0149-4
DOI :
10.1109/CVPR.1999.786934
