Quick Use

We have experimented with some state-of-the-art methods on our dataset. And some convenient and useful tools and SDK are provided for using the dataset.
If you are seeking to do the same, please check out the following to get the work done quickly.

FAST-LIO

FAST-LIO for USTC FLICAR: https://github.com/ustc-flicar/ustcflicar-FAST-LIO
Credit: Forked from https://github.com/hku-mars/FAST_LIO

Left:FAST-LIO with Velodyne HDL32E data and Xsens MTi-G-710 data from HF003 sequence
Right: FAST-LIO with LiVOX Avia(Lidar and imu) data from HF003 sequence

VINS-Mono / VINS-Fusion

When we collect data, we start the sensors in order. Before starting a sensor, program need to check whether the previous sensor is working properly. So at the beginning of the rosbag, not all sensors have finished starting. When using a sensor fusion algorithm (such as camera-IMU), you can play rosbag for a period of time (about 10s), and then start the algorithm after all the sensors are started to prevent initialization errors.

VINS-Mono for USTC FLICAR: https://github.com/ustc-flicar/ustcflicar-VINS-Mono
Credit: Forked from https://github.com/HKUST-Aerial-Robotics/VINS-Mono

VINS-Fusion for USTC FLICAR: https://github.com/ustc-flicar/ustcflicar-VINS-Fusion
Credit: Forked from https://github.com/HKUST-Aerial-Robotics/VINS-Fusion

Left:VINS-Mono with Hikvision1 data and Xsens MTi-G-710 data from HF003 sequence
Right: VINS-Fusion with Bumblebee-xb3-Left/Right data and Xsens MTi-G-710 data from HF003 sequence

ORB-SLAM3

We found that different versions of ORB SLAM3 codes have different configuration requirements and operating effects in different system environments. We provide two versions of ORB_SLAM3 codes that adapted to our experimental environment (Ubuntu 18.04, ROS melodic).

ORB-SLAM3 for USTC FLICAR (IMU): https://github.com/ustc-flicar/ustcflicar-ORB-SLAM3-IMU
ORB-SLAM3 for USTC FLICAR (no IMU): https://github.com/ustc-flicar/ustcflicar-ORB-SLAM3-no-IMU
Credit: Forked from https://github.com/UZ-SLAMLab/ORB_SLAM3

ORB-SLAM3 with Hikvision1 data and Xsens MTi-G-710 data from HF003 sequence

A-LOAM

A-LOAM for USTC FLICAR: https://github.com/ustc-flicar/ustcflicar-A-LOAM
Credit: Forked from https://github.com/HKUST-Aerial-Robotics/A-LOAM

A-LOAM with horizontal Velodyne HDL-32E data from HF003 sequence

A-LOAM with vertical Velodyne VLP-32C data from HF003 sequence

LeGO-LOAM

LeGO-LOAM for USTC FLICAR: https://github.com/ustc-flicar/ustcflicar-LeGO-LOAM
Credit: Forked from https://github.com/RobustFieldAutonomyLab/LeGO-LOAM

LeGO-LOAM with horizontal Velodyne HDL-32E data from HF003 sequence

LeGO-LOAM with vertical Velodyne VLP-32C data from HF003 sequence

LIO-SAM

LIO-SAM for USTC FLICAR: https://github.com/ustc-flicar/ustcflicar-LIO-SAM
Credit: Forked from https://github.com/TixiaoShan/LIO-SAM

Evaluation:EVO

EVO Tools:https://github.com/MichaelGrupp/evo

Overview