Motion Segmentation Using Spanning Trees and Graph Cuts
Abhishek Ranjan · 2003
Motion segmentation in videos involves identifying regions in the frames of the video that correspond to independently moving objects. This is one of the key techniques that help solve various problems encountered when dealing with image sequences, such as redundancy elimination in digital video and tracking of moving objects. Graph theoretic approaches have been widely used for performing segmentation. In this report we have studied some such interesting graph algorithms. We propose an algorithm for segmentation which uses the minimum spanning tree algorithm and a special graph cut, called normalized cut.