Image segmentation of BSE well identifies black precipitates. However, small dark gray features with a similar size of black precipitates has not selected yet.
More SEM work is needed to check:
If the dark gray features are real precipitates. By collecting SE and comparing it with BSE images, it would be possible to see if the contrast is a result of different phases or surface roughness. Other way is to collect EDS data from the same region
The next question is: if the dark gray features are precipitates, do they have the same composition as black ones or they show up in dark gray color because they are not at the surface and located underneath.
Image segmentation of BSE well identifies black precipitates. However, small dark gray features with a similar size of black precipitates has not selected yet. More SEM work is needed to check: