Sparse Motion Segmentation using Propagation of Feature Labels

The paper considers the problem of extracting background and foreground motions from image sequences based on the estimated displacements of a small set of image blocks. As a novelty, the uncertainty of local motion estimates is analyzed and exploited in the fitting of parametric object motion models which is done within a competitive framework. Prediction of patch labels is based on the temporal propagation of labeling information from seed points in spatial proximity. Estimates of local displacements are then used to predict the object motions which provide a starting point for iterative refinement. Experiments with both synthesized and real image sequences show the potential of the approach as a tool for tracking based online motion segmentation.