Lidar Detection and Tracking

Use Kalman Filter and Hungarian algorithm to do tracking work based on 2D Lidar dataset of KITTI dataset.

Demo

screencast-from-18-02-2024-00-20-27_da0IRmyh.mp4

Docker setting

Run

docker run --rm -it --privileged --net=host --ipc=host --gpus all -v /home/jiayansong/workspace/Multi-Object-tracking-2D:/home/venti/Multi-Object-tracking-2D -v /tmp/.X11-unix:/tmp/.X11-unix -e DISPLAY=$DISPLAY -v $HOME/.Xauthority:/home/venti/.Xauthority -e XAUTHORITY=/home/venti/.Xauthority -e ROS_IP=127.0.0.1 -v /home/jiayansong/venti_shared_data:/home/venti/shared_data -v /home/jiayansong/venti_shared_data/nuscenes:/home/venti/nuscenes -v /home/jiayansong/venti_shared_data/nuScenes_rosbags:/home/venti/nuScenes_rosbags --name venti-mot venti-mot

Password

password

Launch ROS

roscore
cd <path-to-nuScenes_rosbags>
rosbag play <rosbag> 

Launch Rviz

roscd detection_and_tracking
rviz -d launch/nuscenes.rviz

Check keyboard input device

cat /proc/bus/input/devicescat /proc/bus/input/devices

Get your keyboard input event (event6 here):

I: Bus=0003 Vendor=046d Product=c341 Version=0111
N: Name="Logitech Mechanical keyboard Logitech Mechanical keyboard"
P: Phys=usb-0000:00:14.0-12.4.4/input0
S: Sysfs=/devices/pci0000:00/0000:00:14.0/usb1/1-12/1-12.4/1-12.4.4/1-12.4.4:1.0/0003:046D:C341.000B/input/input28
U: Uniq=KG511U00000A
H: Handlers=sysrq kbd event6 leds 
B: PROP=0
B: EV=120013
B: KEY=1000000000007 ff9f207ac14057ff febeffdfffefffff fffffffffffffffe
B: MSC=10

Then change the device access

sudo chmod 777 /dev/input/event6