Super-resolution imaging: MCMC and state-space methods

Abstract

Super-resolution (SR) imaging is to produce one or a set of high-resolution images from their lower-resolution counterparts. In this paper, our two recent research works are briefly introduced, that is, the Markov chain Monte Carlo (MCMC) SR approach and the state-space SR approach. The MCMC SR approach exploits the MCMC technique for performing the stochastic SR image reconstruction using image samples that are generated according to the information provided by the observed low-resolution images and the prior knowledge of the unknown high-resolution image. On the other hand, a three- equation-based state-space model is developed by inserting an extra observation equation into the framework of the conventional two-equation-basedKalman filtering to exploit the information from the previously-reconstructed high-resolution frame, the currently-observed low-resolution frame as well as the previouslyobserved low-resolution frame for sequentially producing the high-resolution frames. Extensive experiments are conducted to demonstrate the superior performance of the developed methods.