27 Sep
2015
27 Sep
'15
9:38 a.m.
Easier said than done but a big data solution could be applied to ocr.
Eventually Google 's big data ocr tech will be made available open source
to read pdfs. Until then high quality bitmaps are the logical way to scan,
if you ask me. Eventually throw all docs into a big data lake and let them
churn in there until statistical ocr is applied, a superior approach vs
font pattern matching alone.
Bill Degnan
twitter: billdeg
vintagecomputer.net
On Sep 27, 2015 12:22 PM, "Pontus Pihlgren" <pontus at update.uu.se>
wrote:
On Sun, Sep 27, 2015 at 04:08:07PM +0200, Johnny
Billquist wrote:
I don't have problems reading the current scans, as such. But when
having ten of these open at the same time, and scrolling through
them, it becomes obvious that the bitmaps are heavy. It can take a
while for the screen to be updated. Not to mention the problems you
sometimes hits with searching...
It seems to me that a better tool could solve the issue. One that
could display the OCR:ed content only and the scanned content
only when desired, for instance when you suspect an error.
Is there such a reader? Is the content organised to make it
possible.
/P