ROS Resources: Documentation | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange
Ask Your Question

How to generate pose graph from LIDAR(Laser) data?

asked 2017-03-29 05:19:39 -0500

ZainMehdi gravatar image

Hi, I have been learning g2o to optimize once I have a graph i.e. back-end during my graph slam algorithm implementation. But the issue is I don't have any idea how to build a pose graph from laser data. Lets assume I have bag file that contains all recorded measurement. Then how should I proceed and use this data to generate nodes and edges for my graph. PS: I have read some methods like ICP and dense scan matching that can be used to add edges (constraints) between successive poses. But there application and implementation is naive to me. Kindly help

edit retag flag offensive close merge delete


For the benefit of the community, this discussion is ongoing in

Saurav Agarwal gravatar image Saurav Agarwal  ( 2017-07-24 11:17:31 -0500 )edit

1 Answer

Sort by ยป oldest newest most voted

answered 2017-09-11 00:35:04 -0500

ZainMehdi gravatar image

In order to build a simple graph with laser data (without loop closing) we need to follow following steps 1) Use scan matcher that will match consecutive scans and provide relative transformation between robots last pose from where the last scan was acquired and robot's new pose. 2) The transformation we obtain is relative measurement and can be used to set edge value in g2o 3) To get absolute pose of robot that you can use as vertex value in g2o you need to keep integrating the poses as they arrive. That will give you current pose of the robot from the starting point. 4) Keep the transformations in mind as you just cant add poses speaking abstractly. You have to use transformations i.e. multiply your pose with transformation matrix to get absolute pose.

This will give you a pose graph that you can further play and optimize using loop closure. For further information you can refer to @saurav's blog link text Or my blog which contains a tutorial on graph slam link text

edit flag offensive delete link more

Question Tools



Asked: 2017-03-29 05:19:39 -0500

Seen: 1,340 times

Last updated: Sep 11 '17