Object recognition algorithm
- Here, I used "py38" as the name of virtual environment
conda create -n py38 python=3.8 anaconda
- Activate your virtual environment
source activate py38
(Not confirmed yet)
python3 -m venv ~/.envs/py38
source ~/.envs/py38/bin/activate
python3 -m pip install --upgrade pip
- Make a new catkin workspace and clone the retinanet_ros package
mkdir py38_ws
cd py38_ws
mkdir src
git clone https://github.com/ssteveminq/retinanet_ros.git
cd retinanet_ros/doc
-
Install required packages using pip install ( should use pip install virtual environment)
-
You should confirm that you are using correct pip
which pip
output should be like "/home/$user_name/anaconda3/envs/$environment_name/bin/pip"
pip install -r requirements.txt (**Anaconda**)
python3 -m pip install -r requirements.txt (**python virtual environement**)
source /opt/ros/melodic/setup.bash
cd py38_ws/
catkin build
- clone and build darknet_ros_msgs package
git clone git@github.com:leggedrobotics/darknet_ros.git
cd darknet_ros/darknet_ros
touch CATKIN_IGNORE
cd ..
catkin build darknet_ros_msgs
- build retinanet_ros package
catkin build retinanet_ros
-
Download pre-trained model to detect tire.
-
(Tire) Go To the following link: https://drive.google.com/file/d/16DihIWDBkDreBPA_yqBTu9E3Gawq0QJ4/view?usp=sharing
-
(Barrel) Go to the following link: https://drive.google.com/drive/folders/1G01Ag1YkqYPmBxdPZ7Sv2JJvzCNyGDax?usp=sharing
-
(Barrel-Updated - 2022.MAR, blur, brightness, size ) https://drive.google.com/drive/folders/1bITuhoaSkbuMDPZube8AQNs2Lp-Azkb3?usp=sharing
-
Download the zip file and extract it into a folder called ~/runs/tire-detector/2021-04-22T11.25.25
-
Change the link information in test.py
-
run the test code
source /py38_ws/devel/setup.bash
rosrun retinanet_ros test.py
- You might have to change the topic name for image topic.
- Alexander Witt
@alexwitt23 - Minkyu Kim
- Ryan Gupta