sanqingqu / ros_lane_departure_warning

In this repository, we provide a vehicle lane departure warning package constructed on ROS.

ros_lane_departure_warning

Introduction

In this repository, we provide a vehicle lane departure warning package constructed on ROS. The vehicle lane detection algorithm is constructed from UFLD (Ultra-Fast-Lane-Detection), we provide a ROS wrapper for this algorithm. In order to deploy practical application based on USB cameras, we also provide the ROS-based usb camera driver, which is borrowed from ros-drivers usb_cam. If this repository is useful for you, please star our repository.

We test our package on Ubuntu-20.04 with ROS-Noetic and PyTorch-1.8 (CUDA enabled.). It should be compatible with other platforms.

Prepare

Before you run this package, you need download the UFLD network pre-trained weight and put it to src/ros_lane_detection/ufld_weight, you can download it from Baidu Wang Pan (pwd: 063c)

Usage

To test the results or develop our own codes, we provide three ROS-bag for you, you can download them form Baidu Wang Pan (pwd: p4tv).

Results illustration

ros_lane_departure_warning's Issues

lane departure warning

what is the command to implement lane departure warning?
ros bag play *.bag python3 lane_depart_detect.py
is that true?
thanks, reply...