Smartphone based mass traffic sign recognition for real-time navigation maps enhancement

In this paper, a mobile device based system for recognizing speed limit and end of restriction traffic signs is proposed in the context of data acquisition for the real-time enhancing of navigation maps. Once recognized, the traffic signs are stored on a cloud server and used for updating the geospatial information of the open source OSM maps. The approach is mainly divided into two stages, namely, i) detection and ii) recognition. The detection is achieved through a boosting classifier, while the recognition is performed via a probabilistic Bayesian inference framework that fuses information delivered by a collection of visual probabilistic filters. The stability of the system is evaluated against a ground truth database.