Hi,
i had developped a python code running trough all my photos to check witch photo has the sharpest data for each part of the panorama using yaw, pitch and roll.
This code is then creating the png mask for each of this photos so that ptgui is using the best source among all the pics.
It's working fine with a single pan and tilt camera.
I now want to use a camera that is always facing the subject (a painting) to get better quality. This means that every picture has a different viewpoint.
Could you help me to figure out how to use the yaw, pitch and roll AND vpx,vpy, vpd,vppan et vptilt to calculate wich area of a picture corresponds to wich area of the stiched panorama ?
There must be a formula but i don't manage to find it out.
thanks,
Cyril