ros_vision_pkg

Installation

prerequisite

sudo apt install ros-[distro]-vision-msgs
sudo apt install ros-[distro]-cv-bridge
sudo apt install ros-[distro]-pcl-conversions
sudo apt install ros-[distro]-pcl-ros
sudo apt install ros-[distro]-roslint
sudo apt install ros-[distro]-image-geometry
sudo apt install ros-[distro]-tf2-sensor-msgs
sudo apt install ros-[distro]-tf2-geometry-msgs

sudo apt install libyaml-cpp-dev
sudo apt install python3-pip
sudo apt install python-pip

pip3 install --upgrade pip
pip3 install torch==1.10.1+cu111 torchvision==0.11.2+cu111 -f https://download.pytorch.org/whl/torch_stable.html
pip3 install -r src/requirements.txt

build package

cd ros_vision_pkg
catkin_make
source devel/setup.bash

lane_detector

1. change file permission

   cd src/lane_detector/src/scripts
   chmod +x detect_node.py
   

2. install python dependencies

   pip install -r requirements.txt 
   

3. checkpoint

• Download checkpoint from Link
• Move file as follows src/lane_detector/src/scripts/snapshots/Legacy/latest.pth. Create folder if needed. and locate checkpoint

4. change shebang open file detect_node.py and change the first line starts with #!.

   #!/home/taehoon1018/anaconda3/envs/inspyrenet/bin/python3 --> [your python path]
   

object_detector

1. change file permission

   cd src/object_detector/src/scripts
   chmod +x detect_node.py
   

2. install python dependencies

   pip install -r requirements.txt 
   

3. change shebang open file detect_node.py and change the first line starts with #!.

   #!/home/taehoon1018/anaconda3/envs/yolov5/bin/python3 --> [your python path]
   

• checkpoints yolov5l traffic_sign

drive_scene_parser

1. change file permission

   cd src/drive_scene_parser/src/scripts
   chmod +x drive_scene_parser.py
   

2. install python dependencies

   pip install -r requirements.txt 
   

3. change shebang open file drive_scene_parser.py and change the first line starts with #!.

   #!/usr/bin/python3 --> [your python path]
   

run nodes

Quick Launch (Carla)

source devel/setup.bash
roslaunch src/Launch/vision_carla.launch

Quick Launch (Imcar)

source devel/setup.bash
roslaunch src/Launch/vision_imcar.launch

Launch separately

lane_detector

roslaunch lane_detector lane_detector.launch image_topic:=[image topic from CARLA]

object_detector

roslaunch object_detector object_detector.launch image_topic:=[image topic from CARLA]

drive_scene_parser

roslaunch drive_scene_parser drive_scene_parser.launch

The output of the drive_scene_parser is /image_parsed