The full automation of object visual recognition is a hard and computationally expensive task, mainly for two reasons: on one hand, it is difficult to extract the necessary distinctive information from the raw data, while, on the other, the obtained representations must have characteristics that make them apt to training adaptive classifiers to perform the recognition task of interest. In this paper we present a new method for representing 2-D images, based on the extraction of a set of fractal features, which exploits the approximation of an image with an Iterated Function System, a technique that is already at the basis of many successful image compression tools. In particular, we show that such features have a high discriminatory power, can be easily extracted in an automatic way from the raw data and can effectively be used to train adaptive classifiers to discriminate between different kinds of objects.

Learning to Classify Images by means of Iterated Function Systems

BALDONI, Matteo;BAROGLIO, Cristina;CAVAGNINO, Davide;EGIDI, Lavinia
1998-01-01

Abstract

The full automation of object visual recognition is a hard and computationally expensive task, mainly for two reasons: on one hand, it is difficult to extract the necessary distinctive information from the raw data, while, on the other, the obtained representations must have characteristics that make them apt to training adaptive classifiers to perform the recognition task of interest. In this paper we present a new method for representing 2-D images, based on the extraction of a set of fractal features, which exploits the approximation of an image with an Iterated Function System, a technique that is already at the basis of many successful image compression tools. In particular, we show that such features have a high discriminatory power, can be easily extracted in an automatic way from the raw data and can effectively be used to train adaptive classifiers to discriminate between different kinds of objects.
1998
Fractals and Beyond: Complexities in the Sciences
World Scientific
173
182
9810235933
Iterated Function Systems; Automatic Image Recognition; Automatic Feature Extraction
M. Baldoni; C. Baroglio; D. Cavagnino; L. Egidi
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/2318/103720
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus ND
  • ???jsp.display-item.citation.isi??? ND
social impact