DMCA
Optimal image scaling using pixel classification (2001)
Cached
Download Links
| Venue: | In IEEE International Conference on Image Processing |
| Citations: | 39 - 2 self |
BibTeX
@INPROCEEDINGS{Atkins01optimalimage,
author = {C. Brian Atkins},
title = {Optimal image scaling using pixel classification},
booktitle = {In IEEE International Conference on Image Processing},
year = {2001},
pages = {864--867}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
We introduce a new approach to optimal image scaling called Resolution Synthesis (RS). In RS, the pixel being interpolated is £rst classi£ed in the context of a window of neighboring pixels; and then the corresponding high-resolution pixels are obtained by £ltering with coef£cients that depend upon the classi£cation. RS is based on a stochastic model explicitly re¤ecting the fact that pixels falls into different classes such as edges of different orientation and smooth textures. We present a simple derivation to show that RS generates the minimum mean-squared error (MMSE) estimate of the high-resolution image, given the low-resolution image. The parameters that specify the stochastic model must be estimated beforehand in a training procedure that we have formulated as an instance of the well-known expectation-maximization (EM) algorithm. We demonstrate that the model parameters generated during the training may be used to obtain superior results even for input images that were not used during the training. 1.
Keyphrases
pixel classification optimal image stochastic model superior result simple derivation new approach different orientation resolution synthesis well-known expectation-maximization model parameter high-resolution pixel optimal image scaling input image smooth texture minimum mean-squared error high-resolution image classi cation different class rst classi ed neighboring pixel coef cients low-resolution image training procedure
