Design of a Parallel Face Detection Algorithm for Distributed Low Cost IP-based Surveillance Systems
Wen-Yew Liang, ChiYu Weng -, Yen‐Lin Chen, Che Wun Chiou · Journal of Convergence Information Technology · 2011
In an environment with high secure requirement, multiple video surveillance devices may be installed in the secure environment. In this paper, a multi-camera surveillance system is built for these environments by a set of low-cost IP-based video surveillance cameras. An embedded distributed computing library, called D2MCE, is used to support the distributed computing. It provides a set of easy-to-use interfaces for the programmers to develop distributed parallel programs in a logically shared memory space. The proposed algorithm consists of four phases, including simultaneous intruder detection (SID), distributed camera selection (DCS), distributed & parallel face detection algorithm (DP-FDA), and automatic CAMShift face tracker (ACAM-FT). The advantage is that the system response time for the face detection will be shortened on the low-cost embedded IP-cameras without the help of expensive server machines. The system has been implemented on a set of the Intel PXA270 embedded platforms. Experiment results show that good performance can be achieved on the system.