Resource extension in visual sensor network for panoramic image generation

A visual sensor network (VSN) is constructed by a set of small and low cost sensor nodes which can produce images or videos from a sensed area. Different solutions exist to prolong the lifetime of such networks by considering the remaining energy of the sensors. In this paper surveillance of air space through a terrestrial VSN with randomly distributed sensors is considered. The goal of the network is to form a panoramic image. Hence, it is necessary that the sensors generate a full coverage of the monitored area. It is shown that only a subset of distributed sensors could form the required panoramic image. For this purpose we describe the steps of selection of sensors, formation of full coverage and generation of panoramic image from the captured data. Two novel methods for sensor selection are proposed. For elongation of network's lifetime, we prove that considering the area covered by each sensor, as well as the overlap between the selected sensors, have better results than only considering the energy of each sensor.