Basic feature extraction of segmented objects

[gusqgm]

Once objects have been segmented, basic features can be extracted from them. As a first step, using regionprops from skimage would be sufficient.

[gusqgm]

Outputs of all extracted features should be saved in the table containing the information about each segmented object.

In case objects have nested segmented object groups (e.g. a tissue composed of cells, nuclei, subcellular structures, all segmented independently), each extracted feature pertaining to each nested level should be saved accordingly.

[gusqgm]

One important point to consider is how scalable and how flexible this can be.

Considering e.g. 500K organoids, each with 300+ cells/nuclei.
How fast can we load / write, and how flexible can we be in picking up specific objects based on filters?
-> for example, from several different wells (coming from different plates/experiments), selecting particular cells based on their feature values.

[jluethi]

Closing this, as it is now covered in #69 (initial feature measurements) and #27 (feature storage & access speed questions)