A Hybrid Adaptive Compressive Sensing Model for Visual Tracking in Wireless Visual Sensor Networks

The employ of Wireless Visual Sensor Networks (WVSNs)
has grown enormously in the last few years and have emerged in distinctive applications. WVSNs-based Surveillance applications are one of the important applications that requires high detection reliability and robust tracking, while minimizing the usage of energy to maximize the lifetime of sensor nodes as visual sensor nodes can be left for months without any human interaction. The constraints of WVSNs such as resource constraints due to limited battery power, memory space and communication bandwidth have brought new WVSNs implementation challenges. Hence, the aim of this paper is to investigate the impact of adaptive compressive Sensing (CS) in designing efficient target detection and tracking techniques, to reduce the size of transmitted data without compromising the tracking performance as well as space and energy constraints. In this paper, a new hybrid adaptive compressive sensing scheme is introduced to dynamically achieve higher compression rates, as different datasets have different sparsity nature that affects the compression. Afterwards, a modified quantized clipped Least Mean square (LMS) adaptive filter is proposed for the tracking model. Experimental results showed that adaptive CS achieved high compression rates reaching 70%, while preserving the detection and tracking accuracy which is measured in terms of mean squared error, peak-signal-to-noise-ratio and tracking trajectory.