EM Clustering of Incomplete Data Applied to Motion Segmentation

Many clustering problems in Computer Vision group data points that are the result of statistical estimation and these data points can have a great amount of uncertainty. Motion segmentation by clustering of optical flow is such an example because very often optical flow cannot be estimated without significant uncertainty. We present a EM based clustering algorithm for incomplete data and we apply it to the problem of motion segmentation. The input to the algorithm are the velocity likelihoods and the number of clusters. The algorithm is mathematically very elegant because it does not impose any constraints on the velocity likelihood thus multi-modal likelihood is modeled without difficulty. Coupled with a sophisticated correlated image noise model, the algorithm can handle substantial deviations from the intensity constancy assumption. Experiments with real image sequences show excellent results.