Vehicle surround sensing based on information fusion of monocular video and digital map

Abstract

In this paper we propose a vehicle surround sensing system that combines information from a video sensor with a digital map sensor. The video sensor detects road signs and lane markers, whereas the digital map contains information on road geometry, topology and additional information attributes. We show both sensors to be complementary to each other and fuse their information to obtain driver assistance functions such as Speed Limit Assistant, Curve Warning System, and Hazard Warning System.