Python program demonstrates hand signs detection and hand tracking. Hopefully this script can help in research and development of assistive technology for Autistic and Deaf-mute people.
Principle. Hand geometry model rather than image set trained one, is projected and correlated with sample Image. Natively accelerated by graphics card without direct programming.
Install dependencies:
pip install opencv-python PyOpenGL PyOpenGL_accelerate pygame numpy imutils Pillow==6.1.0
Clone and run:
git clone https://github.com/mmilann/hand_project.git
cd hand_project
python hand_project.py
Tune hand model with input arguments:
python hand_project.py [lenRatio widthRatio thickRatio]
- lenRatio: fingers_length/palm_length ratio
- widthRatio: palm_width/palm_length ratio
- thickRatio: fingers_thickness/palm_thickness
- example:
python hand_project.py 1.05 0.8 0.95