Detecting re-entry of a moving object in an irregular space

Detecting re-entry of a moving object inside an irregular region is one of the most fundamental challenges of Wireless Sensor Networks. The performance of an effective and efficient sensor network for detecting re-entry of moving object is highly related to the proper configuration and design of the network and effective communication between the sensors deployed. For an irregular space of random shape and size, it is practically very difficult to detect an objectpsilas trajectory within the sensing boundary since boundary is not regular. In this paper we have focused on the conversion of a physically random space (concave) to a logically convex hull so that sensing boundary is well defined and then we have detected the network cuts and pockets within that convex hull to detect the status of the moving object. Here we have adopted the technique of using multiple time consecutive frames from two or more sensors currently sensing the object and exploits the time intensity variations with their geometric diversity among the reentry object and their observer sensors.