YOLO + Python + ROS

How to run

• do not have to download darknet, only use opencv

1. download package in your catkin_ws directory
2. catkin_make
3. publish your own image with the topic 'camear/image_color' or just you can modify the topic in yolo.py
4. rosrun ${your package name} yolo.py
5. rosrun image_view image_view image:=MyDarknet/image_annotated

This code is to run YOLO with python + opencv3.3.1 on the ROS(Robotic Operating System) If you publish your own image with the topic "camera/image_color, then this package would re-publish the image-annotated with the topic "MyDarknet/img_annotated"

Reference Paper : Redmon, Joseph, et al. "You only look once: Unified, real-time object detection." Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2016. Site : https://pjreddie.com/darknet/yolo/

Used Opencv Github : https://github.com/opencv/opencv/blob/master/samples/dnn/yolo_object_detection.cpp

Dependency

1. Opencv 3.3.1
2. ROS-lunar
3. python (version does not matter)