On Wed, Sep 14, 2016 at 02:16:38PM +0000, Erika Callagon wrote:
> Thank you for the suggestion, Luis. For one set of data, I have about 30
> images, and I'd like to track a couple thousand of pits. I'm wondering if
> there's a way to automatically track the pits (e.g., by indexing the pits
> based on their locations) per image, and then know which pits are newly
> formed, expanded, or disappeared in each of the images. Or, can this be
> written as a module?

Depending on what the images look like there are ways to cross-identify
the pits once they are marked in individual images.  If the drift is
small then simple methods using centre distances and areas of
intersections for consecutive images could work fine.  But there is no
ready-to-use module for that.  The fundamental missing piece in Gwyddion
is the cross-identification – which would output something like a table
specifying grain numbers in each image that represent the same feature.
Combined with raw grain property dumps (as from Grain Distributions raw
data) it gives all the necesssary information – but some scripting or
external processing will be required anyway.  I might try to write the
identification part as a module that takes a stack of marked images and
dumps a table.



Gwyddion-users mailing list

Reply via email to