Pattern detection in images using LBP-based relational operators

Abstract

This paper describes two new pattern detection image operators, and , called, in a generic way, LBP-based relational operators (LBP-RO). The former is rotational invariant and allows searching for a particular pattern disposes in any direction, the later is a binary operator designed to find image patterns that can be modeled by a pattern function. Both of them are invariants against any monotonic transformation of the image gray scale. We have applied these operators in a case study dedicated to segment the ONH in eye fundus color photographic images. The new segmentation method, called GA+LBP-RO, was compared to a competitive ONH segmentation method in the literature and the results obtained by our method proved to be equal to or better. © 2013 Springer-Verlag.