Hierarchical keyframe-based video summarization using QR-decomposition and modified k-means clustering

摘要: We propose a novel hierarchical keyframe-based video summarization system using QR-decomposition. Specially, we attend to the challenges of defining some measures detect dynamicity shot and extracting appropriate keyframes that assure purity summary. derive efficient compute shots QR-decomposition, utilize it in detecting number must be selected from each shot. Also, theorem illustrates an important property order summarize with low redundancy. The proposed algorithm is implemented evaluated on TRECVID 2006 benchmark platform. Compared other algorithms, our results are among best.