Video streaming in wireless sensor networks with low-complexity change detection enforcement

Video streaming applications in wireless sensor networks are attractive for enabling new pervasive high value services. The reduced amount of available bandwidth in such networks imposes the transmission of small size video frames at very low frame rates. Moreover, when compression techniques cannot be applied, due to constrained computational capabilities of sensor nodes, and uncompressed video frames must be sent through the network, lower image sizes and frame rates must be adopted. In such a scenario, when higher frame rates are required or multiple video streams must be enabled the required transmission bandwidth of each stream must be reduced. In this paper we present a low-complexity algorithm based on change detection which reduces the transmission bandwidth of a video stream based on raw images. Performance results show how the proposed technique can reduce the transmission bandwidth up to 87.22% and 66.78%, with respect to a flat video stream transmission, when low and high changes are experienced in the monitored scene.