[ptx] Re: Enblend suggestion: focus score weighting

[Andrew C Mihal]

On Wed, 16 Jun 2004, Sebastian Nowozin wrote:

>   At first I thought maybe enblend could help and calculate some "focus
> score" value for each image or even just parts of it, and use it to
> prioritize the order of images, leading to an optimal end result, using
> the minimum out-of-focus part of the images possible. And also, as far
> as I understand for enblend an image is either completely redundant
> (removed), or used completely.

You are mostly correct - enblend either uses a pixel completely or not at
all.  This is determined by whether the pixel is inside the input image's
alpha mask or outside. If a connected region of pixels is completely 
overlapped by previous pixels then that region is redundant.

A focus-composition tool could be made as an intermediate tool between 
hugin and enblend. It would adjust the alpha masks to "erase" pixels that 
are deemed out of focus by some metric. The result that enblend would see 
is a set of images with very irregular overlapping regions, but this 
should not be a problem.

Another interesting metric would be: this pixel is too far out of 
consensus with other overlapping input pixels. This might be a way to 
identify ghosts. You have to decide whether to keep the ghost or delete it 
though.

Andrew

---------------------------------
Andrew Mihal
www-cad.eecs.berkeley.edu/~mihal
mihal at eecs.berkeley.edu