Hierarchical approach to fingerprint of places based real-time topological mapping for autonomous rovers Online publication date: Sat, 03-Oct-2015
by Prabhakar Mishra; J.K. Kishore; Harish S. Bharadwaj
International Journal of Computational Vision and Robotics (IJCVR), Vol. 5, No. 4, 2015
Abstract: Autonomous navigation in partially known or unknown environment requires the ability of in motion mapping and localisation. Exploratory rovers employ a hybrid mapping approach comprising of locally metric and globally topological maps to meet resource constraints imposed by the payload. In this paper, we discuss a hierarchical fingerprint of places based approach to generating topological maps using a single camera mounted on the rover. We propose a novel feature extraction process for generation of the finger prints in a monocular vision framework. Each place is characterised by a multi-level hierarchical fingerprint which provides better localisation and loop closure detection. We use a two-pronged strategy for evaluating the performance of this approach: firstly, using standard datasets provided by different universities and secondly, in-situ validation on our experimental rover in indoor and outdoor environments. The proposed algorithm is amenable to real-time implementation in resource constrained rover systems. The comparison of the proposed method with those in literature brings out the computational attractiveness for real-time implementation.
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Computational Vision and Robotics (IJCVR):
Login with your Inderscience username and password:
Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.
If you still need assistance, please email subs@inderscience.com