Python scripts to match up images of pennies from 2 cameras
DataSet is on www.GemHunt.com
Two cameras are pointed at each other. Each one images one side of the coin.
The coin centers need to be correlated to each other:
- The X is mirrored
- They are twisted to each other at some point at some angle.
- The cameras are at different zooms
- The centers are offset
- They might not be parallel to each other.
- The cameras are not calibrated
Instead of trying to do the math for all that, I just use linear regression to match up the centers. It works good!