zhanghanduo / cubicle_detect

obstacle detection based on stereo and YOLOv3
BSD 3-Clause "New" or "Revised" License
0 stars 0 forks source link

Could not find a package "obstacle_msgs" #1

Open zssjh opened 6 years ago

zssjh commented 6 years ago

I find the obstacle_msgs in your repo, but i can not find VisionWorks....

zhanghanduo commented 6 years ago

Sorry that is my mistake. This repository is simply for my personal use, so I did not fine tune for public use. Actually I just updated the readme and you can review it again. There are several things to explain:

  1. This repo is for stereo image input. If you just simply want to use ROS and YOLO, just go to https://github.com/leggedrobotics/darknet_ros with complete solution. Also you can try to use python binding tool (pybind11 to integrate your c++ project with the model built in tensorflow or pytorch.

  2. You can download the obstacle_msgs in my repo. It is simply a customized obstacle msgs package in ROS for 2d and 3d bounding box. You are welcomed to modify it.

  3. The submodule darknet lies in the branch speed which actually derives from the forked darknet. This repo is updating actively and I recommend to you.

You may still encounter some error because I have not extensively tested it. So feel free to raise issues.

zhanghanduo commented 6 years ago

VisionWorks has been deleted in my latest commit. Sorry for wasting your time.