Inspiration
The inspiration is to use projector camera systems to change project dynamically accurate images onto reality.
What it does
The camera sees the projected images, and can analyze the colours and object coordinates. Geometric and radiometric calibration is necessary.
How we built it
I used python library "Dlib" to get 68 facial points, CV2 to control window, coordination and colour manipulation.
Challenges we ran into
Alignment of the pro-cam system. Accessing drives with WSL; WSL is sandboxed so this type of work should be done on the native OS.
Accomplishments that we're proud of
Facial recognition, camera and projector alignment done in time.
What we learned
Manual calibration is somewhat easy.
What's next for AuraPaint
Automatic geometric calibration. Radiometric compensation.
Built With
- camera
- cv2
- dlib
- projector
- python
Log in or sign up for Devpost to join the conversation.