Wireless Camera Network for Image Superresolution

A system to produce a high-resolution image estimate from multiple wireless camera nodes using off-the-shelf components is presented. Wireless camera nodes connected in the network communicate with the vision server through Bluetooth, and are equipped with a camera for image acquisition. The images captured are transmitted to the central vision server. The vision server is responsible for registering the images onto a common projective plane as well as superresolving the low-resolution image set. The RANdom SAmple Consensus (RANSAC) algorithm is used to register the images. The superresolution process used is the Projection Onto Convex Sets (POCS) technique. Reconstruction results are presented.