Estimation of the robot path and mapping of the surrounding environment are some of the most important tasks for a planetary exploration vehicle. Probabilistic SLAM algorithms fulfill these tasks improving the accuracy of the estimations the more and more the same landscape is observed. In this paper two algorithms based on stereo vision and monocular vision are compared on multiple datasets to prove the feasibility of the use of a single camera giving the benefits of mass and complexity reduction of an exploration vehicle. Monocular vision is also suitable to be implemented in UAV applications for planetary exploration purposes. It is investigated the benefit provided by observing an environment rotating the view direction from the direction of forward motion highlighting the best and worst case scenario for monocular vision compared to stereo vision.
A comparison of monocular and stereo visual FastSLAM implementations
2016-06-01
3360310 byte
Conference paper
Electronic Resource
English
Robust stereo visual odometry from monocular techniques
IEEE | 2015
|Convergence and Consistency Analysis for FastSLAM
British Library Conference Proceedings | 2009
|FASTSLAM FILTER IMPLEMENTATION FOR INDOOR AUTONOMOUS ROBOT
British Library Conference Proceedings | 2016
|Convergence and consistency analysis for FastSLAM
IEEE | 2009
|