Open-KTH is a subproject of the AD-EYE testbed for Automated Driving and Intelligent Transportation Systems.
Learn more at https://www.adeye.se/open-kth or contact adeye@md.kth.se
The AD-EYE simulation platform is based on the description presented in the work :
Naveen Mohan, Martin Törngren, "AD-EYE: A Co-Simulation Platform for Early Verification of Functional Safety Concepts", SAE Technical Paper 19AE-0203/2019-01-0126, https://doi.org/10.4271/2019-01-0126
Preprint available at: https://arxiv.org/abs/1912.00448
Data collected by students of the group "AD-EYE on Road" in the capstone course of the Mechatronics Masters program using the AD-EYE platform.
Mapping using the collected data was performed by Maxime Sainte Catherine from the AD-EYE team.
The commands bash extract_pcd_map.bash and bash extract_las_map.bash respectively extract the pcd and the las files.
The data was collected under the form of rosbags using a Velodyne VLP-16 Lidar and a Xsens MTi-700 IMU sensor. Those sensors were mounted on a Renault Twizy which was driven around the KTH campus.
Running the command
bash extract_rosbags.bash
in a terminal will extract the rosbag files.
The ROS package ndt_mapping was used to generate the pcd map using the following settings:
| Setting | Value |
|---|---|
| resolution | 1.0 |
| step_size | 0.1 |
| trans_epsilon | 0.01 |
| max_iterations | 50 |
| leaf_size | 2.0 |
| min_scan_range | 1.5 |
| max_scan_range | 75.0 |
| min_add_scan_shift | 2.0 |
| /approximate_ndt_mapping/imu_topic | /imu/data |
| /approximate_ndt_mapping/imu_upside_down | True |
| /approximate_ndt_mapping/incremental_voxel_update | True |
| /approximate_ndt_mapping/method_type | False |
| /approximate_ndt_mapping/use_imu | True |
| /approximate_ndt_mapping/use_odom | False |