Unified Inverse Depth Parametrization for Monocular SLAM

J. Montiel, J. Civera, A. Davison
2006 Robotics: Science and Systems II  
We present a new parametrization for point features within monocular SLAM which permits efficient and accurate representation of uncertainty during undelayed initialisation and beyond, all within the standard EKF (Extended Kalman Filter). The key concept is direct parametrization of the inverse depth of features relative to the camera locations from which they were first viewed, which produces measurement equations with a high degree of linearity. Importantly, our parametrization can cope with
more » ... eatures over a huge range of depths, even those which are so far from the camera that they present little parallax during motion -maintaining sufficient representative uncertainty that these points retain the opportunity to 'come in' smoothly from infinity if the camera makes larger movements. Feature initialization is undelayed in the sense that even distant features are immediately used to improve camera motion estimates, acting initially as bearing references but not permanently labelled as such. The inverse depth parametrization remains well behaved for features at all stages of SLAM processing, but has the drawback in computational terms that each point is represented by a six dimensional state vector as opposed to the standard three of a Euclidean XYZ representation. We show that once the depth estimate of a feature is sufficiently accurate, its representation can safely be converted to the Euclidean XYZ form, and propose a linearity index which allows automatic detection and conversion to maintain maximum efficiency -only low parallax features need be maintained in inverse depth form for long periods. We present a real-time implementation at 30Hz where the parametrization is validated in a fully automatic 3D SLAM system featuring a hand-held single camera with no additional sensing. Experiments show robust operation in challenging indoor and outdoor environments with very large ranges of scene depth, varied motion and also real-time 360 • loop closing.
doi:10.15607/rss.2006.ii.011 dblp:conf/rss/MontielCD06 fatcat:ssbfq3lkezdspelzabkfsnqbg4