Cellular neural networks for motion estimation

The cellular neural networks (CNN) model is now a paradigm of cellular analogue programmable multidimensional processor array with distributed local logic and memory. CNNs consist of many parallel analogue processors computing in realtime. One desirable feature is that these processors arranged in a two dimensional grid only have local connections, which lend themselves easily to VLSI implementations. We present a new algorithm for motion estimation using a CNN. We start from a mathematical viewpoint (i.e., statistical regularisation based on a Markov random field) and proceed by mapping the algorithm onto a cellular neural network. Because of the temporal dynamics inherent in the cells of the CNN it is well suited to processing time-varying images. A robust motion estimation algorithm is achieved by using a spatio-temporal neighbourhood for modelling pixel interactions