Video Shot Boundary Detection Using Weighted Color Histogram and Gaussian Transition Detection

摘要

In this paper, we propose a novel shot boundary detection algorithm using weighted color histogram. Specifically, we attend to the challenges of detecting gradual shots and extracting appropriate spatio-temporal features that affect the ability of algorithms to efficiently detect shot boundaries. We divide each frame into 9 blocks and utilize their color histogram. Also, we obtain a weight for each block. As the importance of block weighted in determining the distance we derive a new distance function using these blocks. The distance function has abrupt changes in hard cut transitions, and semi-Triangular behavior in gradual transitions. The algorithm detects these transitions by analyzing the distance function. Finally, we will report the results of the experiments using large-scale test sets provided by the TRECVID 2006, which has assessments for hard cut and gradual shot boundary detection. These results confirm the high performance of the proposed algorithm.