A combination of the image sensor with wireless sensor network (WSN) technology has resulted in a new technology, named visual wireless sensor network (VSV). V can be seen as an extension of WSN that image sensor is replaced with scalar sensor. The use of image sensors in visual wireless sensor network brings with it a set of practical and research challenges, so traditional wirele sensor networks approaches are not feasible for these networks. Coverage is one of the fundamental problems in V that directly impact on final quality of the application. In the coverage issue, how the monitored field is sensed with wireless nodes and how the redundant nodes can be used to prolong the network’s lifetime, should be considered. Conventional studies of coverage problem focus on two-dimension or plane coverage. However these simplified plane coverage models cannot accurately characterize the actual situation for visual sensor networks. Hence, accurate monitoring needs algorithms which guarantee full volumetric coverage. So in this thesis, we consider volumetric coverage, in volumetric coverage the aim is to select sensors for cover the monitoring field from minimum height to maximum height. One of the main differences between V and W is sensing model that is an effective factor in coverage problem. So in this thesis sensing model of a sensor is considered as a 3D cone and in order to find sensing neighbors, a method based on intersection of the corresponding cones of each camera has been proposed. In this thesis, visual wireless sensor network is modeled as a hypergraph and coverage problem is reduced to find a minimum hitting set in the hypergraph.Volumetriccoverage problem has been studied in network with randomly and deterministic sensor deployment. In network with randomly deployment, the goal of selection is to obtain maximum coverage and efficient use of network energy. Simulation results show that, estimatevolumetriccoverage with multi planevolumetriccoverage method, that select sensor to coverage multi plane and cluster basedvolumetriccoverage method that, clusters sensor and using formed clusters for selecting sensors to preserve volumetriccoverage have less complexity and is efficient.The coverage rate of these method is closed to greedy method. In greedy method sensors are selected according to accurate information fromthe environment. In networks with deterministic deployment, sensors havethe capabilityof focal length adjustment and regiohadifferent resolution requirement for coverage. Therefore, in this thesis, the objective is selecting cameras and setting the camera’s resolution to fulfill the quality requirements. Keywords: Wireless Visual sensor network, sensing model, volumetric coverage problem, camera selection.
