Abstract | ||
---|---|---|
We present a graph-based SLAM approach, using monocular vision and odometry, designed to operate on computationally constrained platforms. When computation and memory are limited, visual tracking becomes difficult or impossible, and map representation and update costs must remain low. Our system constructs a map of structured views using only weak temporal assumptions, and performs recognition and relative pose estimation over the set of views. Visual observations are fused with differential sensors in an incrementally optimized graph representation. Using variable elimination and constraint pruning, the graph complexity and storage is kept linear in explored space rather than in time. We evaluate performance on sequences with ground truth, and also compare to a standard graph SLAM approach. |
Year | DOI | Venue |
---|---|---|
2010 | 10.1109/IROS.2010.5649205 | IROS |
Keywords | Field | DocType |
distance measurement,monocular vision,visual tracking,map representation,pose estimation,variable elimination,slam (robots),graph theory,relative pose estimation,differential sensor,graph based slam,graph complexity,odometry,robot vision,constraint pruning,complexity reduction,optimization,databases,ground truth,simultaneous localization and mapping,graph representation,visualization | Graph theory,Monocular vision,Computer vision,Variable elimination,Computer science,Odometry,Pose,Reduction (complexity),Artificial intelligence,Simultaneous localization and mapping,Graph (abstract data type) | Conference |
ISSN | ISBN | Citations |
2153-0858 | 978-1-4244-6674-0 | 25 |
PageRank | References | Authors |
0.93 | 11 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Ethan Eade | 1 | 350 | 24.22 |
Philip Fong | 2 | 54 | 3.09 |
Mario E. Munich | 3 | 243 | 18.14 |