The LiDAR tracking library, for tracking objects obtaining from segmentation-based detection and improve segmentation.
## How to use 1. We name your ros workspace as `CATKIN_WS` and `git clone` as a ROS package. ```bash# we recommand you to organize your workspace as following
$ mkdir -p ~/tracking_ws/src
# git clone basic common libraries
$ cd ~/tracking_ws/src
$ git clone https://github.com/xiaxin2000/aeng_563_detection_tracking.git
# build your ros workspace for our Tracking-help segmentation demo
$ cd ~/tracking_ws
$ catkin_make
- Run demo with kitti rosbag:
-
Terminal 1: rosbag play kitti_2011_09_26_drive_0017_synced.bag (https://drive.google.com/file/d/1yJ7xFdPbLgASEE47FuTSNHeCqHDLvpqe/view?usp=sharing)
-
Play rosbag and modify the
detection.yamlandtracking.yamlif necessary. -
Terminal 2: launch Tracking-help Segmentation demo.
$ cd ~/tracking_ws $ source devel/setup.bash $ roslaunch tracking_lib demo.launch
-
detection.yaml and tracking.yaml configure the detection_node and tracking_node in sample. kitti/*.yaml configure the algorithm parameters for KiTTI Dataset, Segmenter.yaml and TrackingWorker.yaml separately for Seg-based Segmentation, Tracking.
./config
├── detection.yaml
├── kitti
│ ├── Segmenter.yaml
│ └── TrackingWorker.yaml
└── tracking.yaml