Hi,
The Junk Dector is great, but sometimes large particles can be incorrectly labeled as junk (with small areas), causing their rejection:
Perhaps having a kind of “dust removal” step - e.g. ignore all defects <0.1% of micrograph area for the purposes of particle rejection - would be useful.
Cheers
Oli
That would be very useful. I’ll need to see whether it junk detects giant viruses! 
Yep, clearly a good idea. I’ll record this feature request. Thanks @olibclarke!
Hi @hsnyder, it is great that this is implemented now, thank you! However, the default is zero - I would suggest maybe setting the default to either the particle area or box area (or some sensible fraction thereof) if an input particle set is provided?
