Visible to the public Biblio

Filters: Keyword is multiple ellipse human model  [Clear All Filters]
2015-05-01
Woon Cho, Abidi, M.A., Kyungwon Jeong, Nahyun Kim, Seungwon Lee, Joonki Paik, Gwang-Gook Lee.  2014.  Object retrieval using scene normalized human model for video surveillance system. Consumer Electronics (ISCE 2014), The 18th IEEE International Symposium on. :1-2.

This paper presents a human model-based feature extraction method for a video surveillance retrieval system. The proposed method extracts, from a normalized scene, object features such as height, speed, and representative color using a simple human model based on multiple-ellipse. Experimental results show that the proposed system can effectively track moving routes of people such as a missing child, an absconder, and a suspect after events.