PCS-TR97-307 Automatic Video Pause Detection Filter.pdf

Dartmouth College Computer Science Technical Report PCS-TR97-307 Automatic Video Pause Detection Filter Xiaowen Liu, Charles B. Owen, Fillia S. Makedon Department of Computer Science Dartmouth College Abstract Increasing interest in multimedia research has been drawn upon the development of video indexing and content-based image retrieval techniques. In this report, we proposed several pause detection algorithms, which instead of searching for signicant visual transitions, the algorithms detect signicant pauses in video streams. A realization of the algorithmswas implemented using ImageTcl toolkit developed at Dartmouth Experimental Visualization Laboratory. In addition to proposing and studying the eectiveness of the pause detection algorithms, another major goal will be to incorporate our algorithms into ImageTcl and test the stability and applicability of the ImageTcl environment. Priliminary experiments showed relatively good results of our pause detection algorithms. 1 Introduction Recently, increasing interest in multimedia research has been drawn upon the development of video indexing and content-based image retrieval. Various schemes for automatic indexing of video databases have been proposed, together with content-based image retrieval techniques. With the increasing signicance of multimedia information, the development of video and image management and retrival system has become more important. While content-based video and image indexing and retrival play an increasingly important role in multimedia information management systems, we propose an alternative development in processing video sequences. Based on the schemes of detecting video cuts, pause detection, on the other hand, is to search for signicant pauses in a given video stream. Video pause detection can be applied to video compression as well as processing of video archives. We propose to design and implement schemes for automatic video pause detection. We will compare dierence schemes pro