期刊名称：International Journal of Advances in Engineering and Management
电子版ISSN：2395-5252
出版年度：2021
卷号：3
期号：2
页码：879-883
语种：English
出版社：IJAEM JOURNAL
摘要：The exploitation of video data requires to extract information at a rather semantic level, and then, methods able to infer “concepts” from low-level video features. This paper adopts a statistical approach and focus on motion information. Because of the diversity of dynamic video content, appropriate motion models are designed and learn them from videos. This paper defines original and parsimonious probabilistic motion models, both for the dominant image motion and the residual image motion. Motion measurements include affine motion models to capture the camera motion, and local motion features for scene motion. The two-step event detection scheme consists in pre-selecting the video segments of potential interest, and then in recognizing the specified events among the preselected segments, the recognition being stated as a classification problem.
