Independent mobility in unfamiliar environments is a significant challenge for people with severe vision impairment. Among other problems, one specific issue concerns the identification of those road signs which can be recognized by sight only. In this paper we present ZebraLocalizer, an application for mobile devices that identifies zebra crossings and guides the user towards them. Two main problems are discussed in this contribution: the identification and localization of the crosswalks, performed by processing data acquired both from the camera and the accelerometers, and the design of an interaction paradigm specifically addressed to blind users. Experimental results, conducted both on a dataset of images and with blind users, validate the applicability of the proposed solution.
ZebraLocalizer: identification and localization of pedestrian crossings / D. Ahmetovic, C. Bernareggi, S. Mascetti - In: Proceedings of the 13th International Conference on Human-Computer Interaction with Mobile Devices and Services : Stockholm, Sweden, August 30-September 02, 2011 / [a cura di] M. Bylund, O. Juhlin, Y. Fernaeus. - New York : ACM, 2011. - ISBN 9781450305419. - pp. 275-284 (( Intervento presentato al 13. convegno International Conference on Human-Computer Interaction with Mobile Devices and Services tenutosi a Stockholm, Sweden nel 2011 [10.1145/2037373.2037415].
ZebraLocalizer: identification and localization of pedestrian crossings
D. AhmetovicPrimo
;C. BernareggiSecondo
;S. MascettiUltimo
2011
Abstract
Independent mobility in unfamiliar environments is a significant challenge for people with severe vision impairment. Among other problems, one specific issue concerns the identification of those road signs which can be recognized by sight only. In this paper we present ZebraLocalizer, an application for mobile devices that identifies zebra crossings and guides the user towards them. Two main problems are discussed in this contribution: the identification and localization of the crosswalks, performed by processing data acquired both from the camera and the accelerometers, and the design of an interaction paradigm specifically addressed to blind users. Experimental results, conducted both on a dataset of images and with blind users, validate the applicability of the proposed solution.
Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
