Rushes Redundancy Detection

rushes is a collection of raw material videos. Various video redundancies exist, such as rainbow screen, clipboard shot, white/back view, and unnecessary re-takes. This pa- per develops a set of solutions to identify and remove these video redundancies as well as create a summary video. We consider the manual editing effects, such as clipboard shots, as a differentiator in the visual language. A rushes video is therefore divided into a group of subsequences, each of which stands for a re-take instance. A rough graphic matching algorithm is developed to estimate the similarity between re-take instances. The experiments on the Rushes 2008 col- lection show that a video can be shortened to 4%-16% of the original size by these redundancy detection solutions. This significantly reduces the complexity in content selection and leads to an effective and efficient video summarisation sys- tem.