2D Pose Graph Optimization

This repository is an implementation of 2D (i.e, 3 degrees of freedom - x,y,theta) pose graph optimization in Python.
A modified gradient descent as authored by Olson et. al. found in refs is implemented.
The code has been tested on the Intel and Manhattan dataset's TORO version.

Odometry motion for Manhattan data

Graph optimization for M3500

Ground truth M3500 data

Graph optimization for Intel data

Ground truth Intel data