Vision-Based Human Motion Analysis for Event Recognition

Event detection and recognition is an active and challenging topic recent in computer vision. The technique could be applied in many visual surveillance systems. This paper describes a new method of recognizing events from video sequences in an office environment. The proposed approach is the vision based human motion analysis via motion history image (MHI) sequences and is invariant to body shape, clothing and position. Combined with edge detection, geometrical characteristics of the human shape in the MHI sequences are considered as the features. An advantage of the proposed method is that the event detection framework is easy to extend by inserting the descriptions of events.