A wavelet-based technique for image refinement

In the paper we concentrate on a Wavelet-based technique of image refinement. Specifically, we are aimed to obtain the extended version of an original image while preserving its quality. The image to be expanded is given in a non-compressed gray-scale format. According to the method we propose, original image is, heuristically, supposed to be a smoothed version of an unknown sharp image we try to reach for. The former will be referred to as the superimage, in the sense that it is not just a resized version of the original, but is rendered in more details than the original. This is the case if we derive the details out of the given image. We have found it reasonable to look at the problem in the Wavelet background, as it comprises the image rescaling and feature extraction operations in their natural combination. We offer a number of approaches to the problem, which are supported by the relevant examples. Applications are, at least, expected to arise in the remote sensing image refinement.