Foveated Imaging systems with space-variant resolution
29 novembre 2018 alle 14.30
Sala riunioni 4° piano - Via Celoria 18 Milano
Relatore: Dott. Matteo Paolo Lanaro
Responsabile scientifico: Alessandro Rizzi
Foveal vision paradigms are traditionally used as ways to achieve information reduction, based on biologic and computational complexity evidence. Such systems compute space-variant images making use of space-variant descriptors especially developed to mimic primate eyes, to decrease the amount of data processed and to reproduce the blurring phenomenon associated with visual acuity. They are mainly used in robotics for embedded machine vision, to create active vision systems with which to study visual attention models, ocular movements and 2-dimensional recognition strategies, like face recognition. In this seminar, after an introduction on space-variant descriptors, their state of the art will be discussed, detailing applications of foveated imaging systems.
Matteo Paolo Lanaro started his PhD student in Computer Science at the University of Milan after graduating in Computer Science for Communications in 2016. He is a member of the MIPS Lab directed by professor Alessandro Rizzi. His research is focused on investigating the sampling of photoreceptors in the human retina, and how their spatial distribution relates to the perception of color. His research fields also include glare, color vision deficiency and spatial color algorithms.