This sample show how to simply use the ZED with OpenPose, the deep learning framework that detects the skeleton from a single 2D image. The 3D information provided by the ZED is used to place the joints in space. The output is a 3D view of the skeletons.
Installation
Openpose
This sample can be put in the folder examples/user_code/OR preferably, compile and install openpose with the cmake and compile this anywhere
The installation process is very easy using cmake.
This sample is a proof of concept and might not be robust to every situation, especially to detect the floor plane if the environment is cluttered.
This sample was only tested on Linux but should be easy to run on Windows.
This sample requires both Openpose and the ZED SDK which are heavily relying on the GPU.
Only the body keypoints are currently used, however we could imagine doing the same for hand and facial keypoints, though the precision required might be a limiting factor.