The authors declare that they have no competing interests.
YSa introduced the hierarchical separation of the feature point trajectories, carried out all the experiments, and wrote the manuscript. YSu introduced the rank estimation of the affine space using the geometric MDL and separation judgement using likelihoods of the fitted affine spaces and revised the manuscript written by YSa. Both authors read and approved the final manuscript.
We propose a new method for segmenting feature point trajectories tracked through a video sequence without assuming a number of independent motions. Our method realizes motion segmentation of feature point trajectories by hierarchically separating the trajectories into two affine spaces in a situation that we do not know the number of independently moving objects. We judge that input trajectories should be separated by comparing the likelihoods computed from those trajectories before/after separation. We also consider integration of the resulting separated trajectories for avoiding too much segmentations. By using real video images, we confirmed the efficiency of our proposed method.