Video segmentation and key frame extraction with parametric model

In this paper, a parametric model for video segmentation and key frame extraction in the video content analysis is proposed. The autoregressive (AR) modeling is used to model the feature sequence of frames over time and to make the future content analysis in the AR parametric space. Based on this parametric framework, detecting shot boundaries in video sequences and extracting key frames from shots are conducted. Real experiments results are presented to illustrate the good performance of this new method.