Dynamic environment modelling and prediction for autonomous systems

Publikation: Beiträge in SammelwerkenAufsätze in KonferenzbändenForschungbegutachtet


This work describes a method to extend classical maps in terms of additional information and a prediction about objects within the environment. The prediction system is based on the behaviour of the observed objects and influences accordingly the updating of the map. First all objects are classified due to their size and further parameters characterizing the ability to move. Furthermore the velocity and orientation of objects within the visible area of the autonomous system are extracted from a vision sensor. In case of a mobile autonomous system they help to adjust its path in real time. Additionally all objects are tracked. In order to generate the statistical map a statistical indicator is introduced describing the possible future positions of objects. Thus the conventional maps can be improved by adding information about the status of the considered space. Furthermore the status of objects can be predicted even when they are not visible anymore. In the case of a mobile system, it will improve the awareness drastically enabling it to act pre-emptively and improve the human-machine interaction in e.g. a production environment.
TitelProceedings of the 2016 13th Workshop on Positioning, Navigation and Communication, WPNC 2016 : WPNC 2016
Anzahl der Seiten6
VerlagIEEE - Institute of Electrical and Electronics Engineers Inc.
ISBN (Print)978-1-5090-5441-1
ISBN (elektronisch)978-1-5090-5440-4
PublikationsstatusErschienen - 17.01.2017
VeranstaltungWorkshop on Positioning, Navigation and Communications - WPNC 2016 - Bremen, Deutschland
Dauer: 19.10.201620.10.2016
Konferenznummer: 13

Zugehörige Projekte

  • Optimierung der Zusammenarbeit von autonomen logistik-Fahrzeugen in einem autonomen cyber-physischem System.

    Projekt: Forschung

Zugehörige Aktivitäten