Energy-aware strategy for collaborative target-detection in wireless multimedia sensor network

Energy-efficiency in visual surveillance is the most important issue for wireless multimedia sensor network (WMSN) due to its energy-constraints. This paper addresses the trade-off between detection-accuracy and power-consumption by presenting an energy-aware scheme for detecting moving target based on clustered WMSN. The contributions of this paper are as follows; 1- An adaptive clustering and nodes activation approach is proposed based on residual energy of detecting nodes and the location of the object at the camera’s field of view (FoV). 2- An effective cooperative features-pyramid construction method for collaborative target identification with low communication cost. 3- An in-network collaboration mechanism for cooperative detection of the target is proposed. The performance of this scheme is evaluated using both standard datasets and personal recorded videos in terms of detection-accuracy and power-consumption. Compared with state-of-the-art methods, our proposed strategy greatly reduces energy-consumption and saves more than 65% of the network-energy. Detection-accuracy rate of our strategy is 11% better than other recent works. We have increased the Precision of classification up to 49% and 65% and the Recall of classification up to 53% and 71% for specific-target and object-type respectively. These results demonstrate the superiority of our scheme over the recent state-of-the-art works.