The GraphSLAM Algorithm with Applications to Large-Scale Mapping of Urban Structures

Abstract¶

This article presents GraphSLAM, a unifying algorithm for the offline SLAM problem. GraphSLAM is closely related to a recent sequence of research papers on applying optimization techniques to SLAM problems. It transforms the SLAM posterior into a graphical network, representing the log-likelihood of the data. It then reduces this graph using variable elimination techniques, arriving at a lower-dimensional problems that is then solved using conventional optimization techniques. As a result, GraphSLAM can generate maps with 108 or more features. The paper discusses a greedy algorithm for data association, and presents results for SLAM in urban environments with occasional GPS measurements.

Summary¶

Formulates full SLAM as a sparse graphical optimization problem over robot poses and map features, then uses variable elimination to reduce and solve the resulting likelihood graph. The paper helped crystallize graph-based SLAM as a unifying alternative to filtering and showed that large urban maps could be handled by exploiting sparsity.

Links¶

Tags¶

Similar Papers¶