Motion segmentation based on factorization method and discriminant criterion

A motion segmentation algorithm based on factor-ization method and discriminant criterion is proposed. This method uses a feature with the most useful sim-ilarities for grouping, selected using motion informa-tion calculated by factorization method and discrimi-nant criterion. A group is extracted based on discrim-inant analysis for the selected feature’s similarities. The same procedure is applied recursively to the re-maining features to extract other groups. This group-ing is robust against noise and outliers because features with no useful information are automatically rejected. Numerical computation is simple and stable. No prior knowledge is needed on the number of objects. Exper-imental results are shown for synthetic data and real image sequences.