Mosaic mode test set

[paul womack]

I have fairly carefully photographed a test set for mosaic mode optimisation.

Every image was taken was a different point, and the tripod was set to two different
heights. 2 Different focal lengths were used. The images are of a road atlas
lying flat.

I attach the files, in screen resolution form (1088x816).
The Hugin project, also included, successfully lines up all 5 images.

I find it interesting (and beyond my understanding) that plane yaw and plane pitch
are NOT required for this set of images.

BugBear

[Bruno Postle]

On Mon, 8 Oct 2018 at 09:23, paul womack wrote:
>
> I have fairly carefully photographed a test set for mosaic mode optimisation.

> I find it interesting (and beyond my understanding) that plane yaw and plane pitch
> are NOT required for this set of images.

Plane yaw and pitch allow you to map to alternative planes!

Imagine that you have photographed a box shaped room with four walls,
a ceiling and a floor - but from various locations and without a
panorama head. In Hugin you can map these images to the six respective
planes and reconstruct a 360° panorama for any possible camera
position in the room. Your road map reconstruction only requires a
single plane, so you only need the default plane (plane yaw and plane
pitch = 0).

--
Bruno

[Roger Broadie]

Hello, Paul.

That's an interesting example. I tried your test case and found, like
you, that it was not necessary to include the plane yaw and pitch
parameters in the optimisation to get a respectable result.

I also tried it on some of my own past examples. I found that,
generally, including the plane yaw and pitch parameters improved the
optimisation ( though not always by much ) but had no significant effect
on the stitch itself. So it's probably better to leave the plane pitch
and yaw parameters out of the optimisation.

Roger Broadie