Event triggered intelligent video recording system using MS-SSIM for smart home security

Abstract

This paper presents an intelligent system for event-triggered video recording for smart home applications. Video recording is triggered through a collaborative sensing strategy. PIR motion detectors are used for both directing the master wireless IP-camera for recording in a specific direction in the entrance hall or initiating other wireless IP-cameras for recording inside the rooms. An activated wireless camera starts video recording only during a targeted motion interval. Motion detection for initiation of the recording process is based on an enhanced Multi-Scale Structural Similarity detection technique. RFID tags are used in all rooms to identify persons entering these rooms. When the moving object shifts to another location at home, the local PIR sends a signal to the Gateway which initiates another video camera. Sensors collaborate for identification of the area to be monitored and the events which are to be recorded. The proposed system helps cover all smart home areas, save the required storage space and speeds-up video event analysis. Keywords: Smart homes, Multi-modal collaborative sensing, Intelligent video recording, Event-triggered recording, Motion detection, Structural similarity index