Jochen Sprickerhof
Last changed: 2012-09-24

Large-Scale 3D Mapping with Explicit Loop Closing

by Jochen Sprickerhof, Andreas Nüchter, Kai Lingemann, and Joachim Hertzberg

Large-Scale 3D Mapping with Explicit Loop Closing

Simultaneous Localization and Mapping (SLAM) is the problem of building a map of an unknown environment by a mobile robot while at the same time navigating the environment, using the unfinished map. For SLAM, two tasks have to be solved: First reliable feature extraction and data association, second the optimal estimation of poses and features. These two parts are often referred to as SLAM frontend and backend. Algorithms that solve SLAM by using laser scans commonly rely on matching closest points in the frontend part. Then the SLAM front- and backend have to be iterated to ensure that the map converges.

This paper presents a novel approach for solving SLAM using 3D laser range scans. We aim at avoiding the iteration between the SLAM front- and backend and propose a novel explicit loop closing heuristic (ELCH). It dissociates the last scan of a sequence of acquired scans, reassociates it to the map, built so far by scan registration, and distributes the difference in the pose error over the SLAM graph. We describe ELCH in the context of SLAM with 3D scans considering 6 DoF. The performance is evaluated using ground truth data of an urban environment.


An animated comparison between ELCH and LUM can be seen in the video. It compares our previous strategy LUM (on the left) with our new strategy ELCH. It shows the different steps during computation es depicted in the Figure above and the elapsed computing time.

Gernerated 3D Map

Here you find more 3D views of the final map.

In the following the left image presents the 3D map before loop closing, while the right shows the corrected, consistent view:

3D view before loop closing3D view after loop closing


We thank Oliver Wulf and Bernardo Wagner (both Leibniz University of Hannover, Germany) for making the 3D data set Hannover2 publicly available and Claus Brenner (Leibniz University of Hannover) for letting us use the airborne 3D data.