Understanding multifractality: reconstructing images from edges
统计力学
2007-05-23 v1 无序系统与神经网络
摘要
It has been recently proven that natural images exhibit scaling properties analogue to those of turbulent flows. These properties allow regarding each image as a multifractal object, for which its most singular manifold conveys the most of the non-redundant structure. In the present work, we go further in this analysis, proposing a simple propagator that reconstructs the whole image from this set. This fact could have deep implications for biology, technology and statistical mechanics.
引用
@article{arxiv.cond-mat/9811372,
title = {Understanding multifractality: reconstructing images from edges},
author = {Antonio Turiel and Angela del Pozo},
journal= {arXiv preprint arXiv:cond-mat/9811372},
year = {2007}
}
备注
Latex, 13 pages, 6 figures formed by 15 PostScript files. Higher quality pictures may be requested from authors