Hands-on Localization

Description

This repo contains the implementation of pose graph SLAM using a 2D lidar and imu. Everything was implemented from scratch including motion model, measurement model, prediction and update. This followed a filtering approach with Extended Kalman Filter being the filter of choice.

Results

Scan alignment using ICP

SLAM

For an extended video click

SLAM

Dependencies:

pip install shapely

pip install scipy

last update:

How to Run on Hardware

1. roslaunch pose-graph-slam test_kobuki_basic.launch
2. rosrun pose-graph-slam working.py
3. roslaunch pose-graph-slam kobuki_keyboard_control1.launch

How to run in simulation

launch sim

1. roslaunch turtlebot_simulation turtlebot_hoi_circuit1.launch or
2. roslaunch turtlebot_simulation turtlebot_hoi_circuit2.launch

Launch manual control nodes

2. roslaunch pose-graph-slam kobuki_keyboard_control.launch

Launch SLAM

4. rosrun pose-graph-slam integration.py