It is just a very blunt approach to background removal!
I am not sure if the option exists in Photoscan Standard but in Pro you can import masks from the existing geometry using Tools->Import->Masks... then select "Active Chunk" and "From Model".
If your background is causing problems with your alignment or reconstruction, or you want to force photoscan to find more matches on your subject (guessing that's how it works) you can do a very quick geometry construction, after aligning images as well as you can, and use that geometry to quickly mask out most of the background.
You may need to scan through the photos after you imported the mask to see if any were incorrectly aligned in which case you will need to edit the mask manually.
Now you can align the images again, using the mask constraint, and the masked out background will no longer be taken into account when aligning.
I haven't done any tests yet to confirm whether this increases workflow speed or quality but it did help me align some photos i took of a friend who couldn't stand still... photoscan was aligning images based on the static background because my friend was swaying too much. quickly masked out the background using this approach and the photos aligned much better on his swaying body and i was able to do a reconstruction, although it was still awful... no substitute for a 50 camera rig!