Omnidirectional Surveillance System Featuring Trespasser and Faint Detection for i-Habitat

摘要

This paper proposed an efficient omnidirectional surveillance system for i-habitat (intelligent habitat or smart home in another name) security. In this surveillance system, the omnidirectional scenes in a room, kitchen, balcony, dining hall, corridor, garage or backyard within an i-habitat are first captured using wireless webcam attached to a specially designed hyperbolic optical mirror. The captured scenes are then fed into a wireless router connected to laptop computer for image processing and alarm purposes. A mapping algorithm is developed to match the captured omnidirectional image into panoramic image, hence providing the observer or image processing tools a complete wide angle of view. An efficient trespasser detection algorithm is designed for trespasser detection purpose, and an efficient faint detection algorithm is designed to identify/monitor if there is any member who stay in i-habitat (old-folks or ill-patients) who fall down or faint and thus, immediate attention can be carried out. Based on the experimental findings, both the proposed trespasser and faint detection algorithm achieves high accuracy.