- Open an Anaconda Prompt (in Windows, or any terminal on Mac/Linux) and enter the following comands
conda create -n freemocap-env python=3.7
conda activate freemocap-env
pip install freemocap -v
ipython
import freemocap as fmc
fmc.RunMe() #this is where the magic happens. Also, as of 2021-07-08, is likely going to be pretty crashy/buggy2021-06-12_FreeMoCap_Clips_16MB.mp4
- Install Anaconda
- Install CUDA
- Install OpenPose (Windows Portable Demo)
- https://github.com/CMU-Perceptual-Computing-Lab/openpose/releases/tag/v1.6.0
- This can be complicated, so I'll need to add more instructions here!
- https://github.com/CMU-Perceptual-Computing-Lab/openpose/releases/tag/v1.6.0
- Install ffmpeg (to save the final animation)
- Two or more USB webcams attached to viable USB ports
- (USB hubs typically don't work)
- Each recording must (for now) start with an unobstructed view of a Charuco board generated with python commands (or equivalent):
import cv2 aruco_dict = cv2.aruco.Dictionary_get(cv2.aruco.DICT_4X4_250) #note `cv2.aruco` can be installed via `pip install opencv-contrib-python` board = cv2.aruco.CharucoBoard_create(7, 5, 1, .8, aruco_dict) charuco_board_image = board.draw((2000,2000)) #`2000` is the resolution of the resulting image. Increase this number if printing a large board (bigger is better! Esp for large spaces! cv2.imwrite('charuco_board_image.png',charuco_board_image)
Follow the GitHub Repository and/or Join the Discord (https://discord.gg/HX7MTprYsK) for updates!