Computer Vision Course Project, WPI

Problem Statement:

Do anything interesting with a Rubik's cube

Content:

1. Detection of Rubik's cube
2. Augmented Reality
3. Pose Estimation

Augmented Reality over the Rubik's cube using Perspective Projection:

Augumented objects from waveform object files to the face of Rubik's cube

Pose Estimation of the camera with respect to Rubik's cube using Perspective-N-Point:

The yaw is very noisy, this is one bug that needs to be fixed. Kalman filter can be used to reduce the estimation noise. The scale has to be found to get accurate state estimation

References

1. Plane Tracking
2. Augumented Reality
3. YCB Dataset