Spatial correlation model for heterogeneous cameras in wireless multimedia sensor networks

Abstract

Collaborative in-network processing has shown to efficiently reduce the per-node processing and transmission requirements in wireless multimedia sensor networks. When the camera nodes are randomly distributed in an area of interest, it is very likely that their fields of view would overlap causing the corresponding visual information to be highly correlated. In applied wireless multimedia sensor networks, heterogeneous camera nodes with different sensing radii and angles of view are usually deployed, which has shown to enhance the overall network performance and lifetime. This paper introduces a new correlation model to exploit the correlation characteristics among heterogeneous camera nodes for wireless multimedia sensor networks. The proposed model, of which a closed-form analytical correlation function was derived, takes into consideration different sensing parameters of the camera nodes; such as the sensing radii and the angles of view. The simulation results demonstrated that the proposed model outperforms the state-of-the-art in terms of the accuracy of estimating the correlation characteristics, the information gain, and the distortion ratio.