Fractal and integral geometry tools for texture deformation measurement
Abstract
In this work, we investigate the use of two techniques for segmentation of different states of one texture (e.g. deformations of an homogeneous texture): - fractal geometry, that deals with the analysis of complex irregular shapes which cannot well be described by the classical Euclidean geometry - Integral geometry, that treats sets globally and allows to introduce robust measures.We focus on the study of two parameters, Lacunarity and Favard length, and proove a theoritical link between them. As an application,we are able to achieve with an excellent accuracy automatic classification of Lung diseases on the basis on SPECT images. Classical techniques tried on those images give poor results.