[campus icon] Accesskey [ h ] University of Paderborn - Home
DE deutsch
Die Universität der Informationsgesellschaft


Robust and Precise 3D SLAM at Low Computational Cost Using Laser Scanner and Camera Data

Bearbeiter:Philip Frieling

Autonomous mobile robot applications require that the environment and the robot pose is precisely known. Currently, the GET Lab uses a height-map approach to map the environment and estimate the robot pose. This restricts the robot to only operate in an environment that can be mapped on a plane surface. To overcome this issue and allow autonomous operations in complex environments, the goal of this thesis is do develop a 3D localization and mapping system, which is designed to run on the rescue robot GETjag. Object detection as well as path planning in multi leveled environments should be supported by creating a colored 3d map of the environment. The new system has to run at real-time, keep resources free for other applications and may only use the existing sensors. Backwards compatibility must be ensured by offering a 2.5D map. For evaluation, the SLAM system will be tested in real world indoor and large urban outdoor environments at the Paderborn University.