A device-independent algorithm for the estimation of an enhanced resolution image from a low-resolution compressed sequence is proposed. The algorithm utilises least squares matching to extract the interdependence of the low-resolution images. This algorithm also has the effect of attenuating compression artefacts. Improving the spatial resolution of the image sequence is not the only goal of the enhancement algorithm, as the enhanced images in turn lead to digital elevation models (DEMs) of improved accuracy. Experimental results illustrate the algorithm's performance as a tool for digital photogrammetric applications. Stereoscopic sets of left and right images were taken of objects of known geometry and DEMs were created using both the original coarse images and compressed images enhanced by the algorithm.
The Photogrammetric Record Vol. 19, Issue 106, p. 149-162