This repository contains an implementation of the robust map merging method presented in
J. G. Mangelson, D. Dominic, R. M. Eustice, and R. Vasudevan,
“Pairwise Consistent Measurement Set Maximization for Robust Multi-robot Map Merging.”, ICRA 2018.
The purpose of this package is to find a subset of inliers in a large set of inter-robot loop closures. Due to large measurement errors and perceptual aliasing in multi-robot applications, we need a robust method to select which measurements to trust.
An example of usage of the libraries is available in the examples/
folder.
catkin build
or catkin_make
.You can also build it normally using cmake.
rosrun robust_multirobot_map_merging robust_multirobot_map_merging_node <trajectory robot1 .g2o file> <trajectory robot2 .g2o file> <inter robot loop closures .g2o file>
_Some .g2o files are available for testing in the pose_graph_datasets/
folder._