Image Representation by Compressive Sensing for Visual Sensor Networks
This paper addresses the image representation problem in visual sensor networks. The authors propose a new image representation method for visual sensor networks based on Compressive Sensing (CS). CS is a new sampling method for sparse signals, which is able to compress the input data in the sampling process. Combining both signal sampling and data compression, CS is more capable of image representation for reducing the computation complexity in image/video encoder in visual sensor networks where computation resource is extremely limited.