Added binary fill to surface labelled droplet reconstruction

[jo-mueller]

Description

This makes the reconstruction workflow robust for surface-labelled droplets. A remaining concern is what happens to partially labelled droplet or a droplet that has a hole in the labelled area. But maybe that's for a future PR to look at.

Type of Change

• Bug fix (non-breaking change which fixes an issue)
• Documentation update

How Has This Been Tested?

• Workflow tested interactively
• Workflow tested from notebook

Checklist:

• My code follows the style guidelines of this project
• I have performed a self-review of my own code
• I have commented on my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• My changes generate no new warnings
• New and existing unit tests pass locally with my changes

[jo-mueller]

@carlos-gomez00 @Schroetinger I did a comparison of the differences in the stresses for the double-labelled droplet. You find all the results in this notebook.

Any comments on something that's missing from the analysis?