Detect a road lane using RGB camera
- 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.
/camera/color/image_raw
([sensor_msgs/Image]) THe camera measurements
-
/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
cd ~/catkin_ws/src/
git clone https://github.com/Thermay-Robotics/turtlebot3_lane_detection.git
cd ~/catkin_ws
catkin_make
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
Ubuntu 20.04 LTS
ROS Noetic
Jetson Nano
Ubuntu 18.04
ROS Melodic