Automatische Erzeugung langzeitverfügbarer Punktmerkmalskarten zur robusten Lokalisierung mit Multi-Kamera-Systemen für automatisierte Fahrzeuge

Abstract

Automated vehicles heavily rely on accurate self-localization to operate safely and efficiently in dynamic road traffic scenarios. In this work, an iterative and robust approach to create highly accurate maps comprising any number of drives of arbitrary length with constant computation complexity is proposed. Furthermore, a high-precision self-localization method using multiple cameras covering the surrounding environment is presented.