turtlebot3_lane_detection

Detect a road lane using RGB camera

Notice

• The node subscribes to the camera topic detect the lane. A region of interest is defined at the bottom of the image. The ROI is used to detect lane. To do so, we detect each sides of the lane and compute a mean lane. Then, we compute the angle between the ordonate axis and the mid-lane. This value is then published.

Subscribed Topics

• /camera/color/image_raw ([sensor_msgs/Image]) THe camera measurements

Published Topics

• /lane_detection/angle ([turtlebot3_lane_detection::line_msg]) Publishes the angle betwenn the lane and the robot

• /lane_detection/image/image_lane ([sensor_msgs/Image]) Publishes the detected lane and the mid lane

How to build

cd ~/catkin_ws/src/
git clone https://github.com/Thermay-Robotics/turtlebot3_lane_detection.git
cd ~/catkin_ws
catkin_make

Run

Launch the camera node (please update the subscribed topic camera name in the launch file)

Launch detection node

roslaunch turtlebot3_lane_detection turtlebot3_lane_detection.launch

if you want to run it with the camera, you can launch

roslaunch turtlebot3_lane_detection turtlebot3_detection_realsense.launch

Test environment

Ubuntu 20.04 LTS
ROS Noetic

Jetson Nano
Ubuntu 18.04 
ROS Melodic