Large-scale image sensing by a group of smart image sensors

Large-scale image sensing over the extended area with multiple image sensors is considered very effective for collecting and analyzing global information, which is hard to be achieved by a single or small number of image sensors. However, if they share the network resources simultaneously, how to transfer and process vast amount of data caused by numerous sensors is one of the critical issues to implement a practical system. In this paper, we present a new approach to large-scale image sensing system using random accessible smart image sensors to reduce the data volume at image acquisition level. By the pixel-level data management throughout the whole processing, both network and system resources can be saved while preserving the main region of the scenes. Based on the simulation results regarding spatial/temporal resolution schemes, an experimental system using multiple sensors is implemented and the basic results are provided.