Vehicle Localization by Utilization of Map-based Outline Information and Grayscale Image Feature Extraction

K. Schönherr, B. Giesler, and A. Knoll (Germany)

Keywords

image processing, pose estimation, object tracking

Abstract

Todays vehicles are more and more equipped with video cameras. These cameras are used e.g. for lane and parking assist systems. For the localization of the vehicle in the world the Global Positioning System is widely used. Unfortunately, GPS is very imprecise and insufficient for many driver assistance applications. To overcome this limitation, a more precise localization using a different approach has to be found. High quality localization sensors are already available on the market but still too expensive for mass integration in automobiles. The aim is to extract additional information from already established car equipment and to use it for precise localization. Our approach combines map information with extracted image features using a model based algorithm.

Important Links:



Go Back