[robotics-worldwide] [Software] Open source visual slam for urban and indoor environments

Dezhen Song dzsong at cs.tamu.edu
Mon Aug 10 13:41:45 PDT 2015

Dear Colleagues,


We are announcing  source code release for  Multilayer Feature Graph (MFG)
based Visual SLAM. Please see 




for details. This effort is part of the following paper appeared at a recent
issue of IEEE Transactions on Robotics:


Yan Lu and Dezhen Song, "Visual Navigation Using Heterogeneous Landmarks and
Unsupervised Geometric Constraints", IEEE Transactions on Robotics (T-RO),
vol. 31, no. 3, June 2015, pp. 736 - 749. See
351> &arnumber=7103351


Modern visual navigation approaches mostly use homogeneous features (e.g.
points) as landmarks. However, this choice cannot fully exploit the
information  in the heterogeneous landmarks from man-made environments.
There exist abundant lines in parallel directions and salient building
facades in typical urban environments. These heterogeneous landmarks and
their inner geometric relationships can help improve navigation performance
if the geometric constraints between them are modeled and utilized properly.
Moreover, points, lines and planes are more robust to lighting changes than
points alone, which can also allow multiple robots to share their knowledge
of space. 


This work has been a joint effort of many researchers including Yan Lu,
Joseph Lee, Madison Treat, Chieh Chou, and many others from Netbot Lab at
Texas A&M, Haifeng Li and Jingtai Liu from Nankai University, China, Yiliang
Xu (Kitware, Apple), A. G. Amitha Perera (Kitware, Google),  and Sang Min Oh
(Kitware, Nvidia).  See credits paper for details.


-Dez Song


Dezhen Song, Ph.D., Associate Professor

Department of Computer Science and Engineering,

311B HRBB, TAMU 3112,

Texas A&M University,

College Station, TX 77843-3112

Phone +1(979)8455464



More information about the robotics-worldwide mailing list