A Novel Recursive Region-Based Camera Break Minimization Algorithm

摘要

This paper presents a novel recursive region-based camera break minimization algorithm to minimize the number of shots to be used for video browsing and searching. Regions are used as basic element to represent shots. The proposed recursive algorithm reduces the number of shots by either eliminating shots or merging shots, based on some pre-defined cases (or rules). Additional cases may easily be included in the algorithm to enhance the algorithm. A cost function is also proposed in solving the problem of selecting shots for merging in complicated situations. Basic parameters, such as the length of the shot and the number of common regions are used in this paper. Any sensitive and meaningful parameters added to the cost function may further improve the accuracy in the process of the shot selection. Our experimental results show that the algorithm is capable of minimizing the number of shots to any required value.