This small toolset allows to integrate SLAM solution provided by MAD-ICP with HDMapping. This repository contains workspace that :
- submodule to tested revision of MAD-ICP
pip install --upgrade pip setuptools wheelpip install mad-icpDownload the dataset from NTU-VIRAL For this example, download eee_03.
rosbags-convert --src {your_downloaded_bag} --dst {desired_destination_for_the_converted_bag}Changes:
Create new config file:
code ~/.local/lib/python3.10/python3.10/site-packages/mad_icp/configurations/datasets/Create a new file named
ntu-viral.cfgAdd the following content:
min_range : 1.3
max_range : 120
sensor_hz : 20
deskew : True
rosbag_topic : "/os1_cloud_node1/points"
lidar_to_base:
- [1, 0, 0, 0]
- [0, 1, 0, 0]
- [0, 0, 1, 0]
- [0, 0, 0, 1]Run MAD-ICP
~/.local/bin/mad_icp --data-path {path_for_rosbag2} --estimate-path {path_for_output} --dataset-config ~/.local/lib/python3.10/site-packages/mad_icp/configurations/datasets/ntu-viral.cfg