Abstract

We present an approach for motion segmentation using in- dependently detected keypoints instead of commonly used tracklets or tra jectories. This allows us to establish correspondences over non- consecutive frames, thus we are able to handle multiple ob ject occlusions consistently. On a frame-to-frame level, we extend the classical split-and- merge algorithm for fast and precise motion segmentation. Globally, we cluster multiple of these segmentations of different time scales with an accurate estimation of the number of motions. On the standard bench- marks, our approach performs best in comparison to all algorithms which are able to handle unconstrained missing data. We further show that it works on benchmark data with more than 98% of the input data miss- ing. Finally, the performance is evaluated on a mobile-phone-recorded sequence with multiple ob jects occluded at the same time.