Regions of interest detectors based on symmetry

Vito Di Ges`u and Cesare Valenti

Dipartimento di Matematica e Applicazioni, Universit`a di Palermo (Italy)

digesu,cvalenti@math.unipa.it

Abstract

This paper briefly reviews some methods to evaluate local and global image symmetries based on gradient and gray level information. The relevance in studying symmetry derives from its ability in detecting regions of interest that are at the basis of further image elaboration. Applications to robotic vision, face analysis and classification and examples on real data are also shown.

Back                                            Home