A Multirate Approach to High-Resolution Image Reconstruction

J.W. Scrofani and C.W. Therrien (USA)


Multirate, Super-resolution, Image Reconstruction, Opti mal Filtering


The paper addresses the problem of reconstructing a high resolution image from a set of observation images sampled at a lower rate and subject to additive noise and distor tion. The method introduced here is based on our work in multirate optimal filtering, extended to two dimensions. The linear filters used for the reconstruction are periodi cally spatially-varying (in 2-D) and chosen so that their re gions of support are closest to the point being estimated. Results are presented for images with additive white noise and compared to methods using interpolation.

Important Links:

Go Back