Conclusion

Abstract

VSNs present a low-cost, low-power visual solution for a wide range of applications ranging from surveillance to telepresence. The network consists of multiple collaborative visual nodes with sensing, data processing, and communication capabilities. This enables them to collect large volumes of images about surveyed scenes, process them, and transmit extracted data to the base station for further analysis. VSNs are faced with several challenges given their limited resources (memory, power, and bandwidth). This book addressed the issues raised in surveillance using VSNs, in particular those related to the underlying image processing steps performed at the camera ends. These include image registration, image fusion, object detection, and object tracking. A survey of these processing steps was provided, which may be beneficial for other researchers in the area. Simple, lightweight, and yet accurate algorithms for smart embedded visual sensing nodes were then presented and discussed. This set of intelligent low-power algorithms extends the reach of video surveillance to a wider range of applications. Critical parts of the algorithms were handled via hardware architecture assists that alleviate the burden of some of the computationally heavy and memory demanding components.