Mobile visual search from dynamic image databases

Abstract

Mobile phones with integrated digital cameras provide new ways to get access to digital information and services. Images taken by the mobile phone camera can be matched to a database of objects or scenes, which enables linking of digital information to the physical world. In this paper, we describe our method for mobile image recognition, which is a part of a pilot system for linking of magazine page images to additional digital content. Such magazine databases are highly dynamic, so the recognition method needs to support addition and deletion of images without rebuilding the whole database. Meanwhile we significantly reduce the memory cost in the system without sacrificing retrieval accuracy. We present recognition results with two different databases. © 2011 Springer-Verlag.