Dear Kasper,
I am glad you enjoyed the course.
Mikael Agn, a former master student at NRU, did a small study in
his master thesis
(
http://www2.imm.dtu.dk/pubdb/views/publication_details.php?id=6647)
in section 5.3 where he investigated manual edits vs. no manual
edits on the same sample.
I have repeated Mikael's analysis on more datasets at NRU from 3
scanners and saw that the difference was scanner dependent -
meaning for one scanner the surface errors were systematic in the
same place due to poor image quality, whereas for others it
wasn't.
If you'd like more details let em know,
cheers,
Melanie
On 28-08-2014 12:54, Kasper Jessen wrote: