A-LOAM-Docker

This repo contains Docker image for A-LOAM and useful information for working with it.

Building and launching Docker image

At first you need to build Docker image.

cd <path-to-this-repo>
docker build -t a-loam:latest .

Then you can run Docker image with our script:

xhost +
./run.sh <path-to-A-LOAM-repo>

You can get A-LOAM here.

Using Docker image

Dataset playing

For playing Velodyne VLP-16 dataset you need to connect to the existing Docker container from the second terminal instance:

docker exec -it <container-id> bash

Then you can play Rosbag:

rosbag play <path-to-dataset>

Rosbag recording

You can record Rosbag for saving dataset playing results from the third terminal instance:

rosbag record /aft_mapped_to_init /laser_cloud_map /velodyne_cloud_registered

NOTE: You can record any other topics, but we will use only these ones.

Getting point clouds

Use the provided notebook

Getting trajectory

Per frame trjaectory is available in /aft_mapped_to_init topic.

Coverting from ros2 to ros1:

Using rosbags python lib

to install it:

pip install rosbags

and then use

rosbags-convert <dir-to-ros2-bag> --include-topic <topic_one --include-topic <topic_two> 
#example 
rosbags-convert data_folder --include-topic /velodyne_points --include-topic /imu

This only works for supported messages

in our case pointclods and IMUs are supoorted

Addtional Method:

The notebook is now updated with method to save selected topics from a ROS2 bag. you can use is to save sub ros2bag and then use rosbags python lib

rosbags-convert <dir-to-ros2-bag>
# Example
rosbags-convert data_folder

rectification matrix in SE3
process map
project LOAM to 2D
save plots for PC map projection by LOAM sequences

python project_map.py --map_path /path-to-map.pcd --bag_path path-to-bag.bag --save_path path-to-save-materials

Plot map in 2D with Localization odometry

given the map and odom file from localization, run the command below to plot solution and map in 2D. we will store:

rectification matrix in SE3
process map
project localization to 2D
save plots for PC map projection by localization sequences

python bag2_2Dmap.py --map_path /path-to-map.pcd --poses path-to-odom.csv --save_path path-to-save-materials

Tags to GNSS transformation

python gnss/tag2gnss.py --map_3d /path-to-map.pcd --tags_num number-of-used-tags --gnss_coords /path-to-stored-tags-GNSS\

NOTE, keep the order while clicking on tags position on the map with the order of GNSS coordinates.

use right click to choose points on the map.

rttariverdi67/A-LOAM-Docker