Face Detection and Tracking in a Wireless Vision Sensor Network Environment

Vol.3 No.4

Year: 2009

Issue: April-June

Title: Face Detection and Tracking in a Wireless Vision Sensor Network Environment 

Author Name: Chia W.C, Yeony L.S, Ngau W.H, Chew L.W, Kenneth Li-minn , Seng K.P 

Synopsis: 

This paper presents a practical implementation of a Wireless Visual Sensor Network (WVSN) with a back-end face detection and tracking system. This back-end processing is based on the method proposed by Viola and Jones. The WVSN is a visually enabled wireless sensor network (WSN). The WVSN consists of visual nodes, network motes, and a base station. The visual nodes are used to capture visual data that is relayed to the base station via network motes or other visual nodes which can act as network motes. The base station acts as a gateway between the WVSN and the back end. To ensure reliable and robust network connectivity, the WVSN is built upon a mesh network allowing automatic signal rerouting if the current path fails. This network is based on the EmberZNet stack. In this implementation, the visual nodes are a combination of a CMOS camera and Xilinx Spartan-3 FPGA upon a wireless ZigBee network using the Ember EM250 System-on-Chip devices.