作者: P. Punitha , Joemon M. Jose
DOI: 10.1007/978-3-642-10543-2_13
关键词:
摘要: Detection of shot boundaries in a video has been an active for quite long time, till the TRECVID community almost declared it as solved problem. A problem is assumed to be when there no significant improvement being achieved from that state-of-the art methodologies. However, certain aspects can still researched and improved. For instance, finding appropriate parameters instead empirical thresholds detect very challenging researched. In this paper, we present fast, adaptive non-parametric approach detecting boundaries. Appearance based model used compute difference between two subsequent frames. These frame distances, are then locate The proposed boundary detection algorithm uses asymmetric region support automatically adapts Experiments have conducted verify effectiveness applicability method segmentation.