actually, if you have OCR done, the rest would be fairly easy. every
puzzle has a grid, which is a perfect set of clues for your algorithm
to realign the sheet.
mike
--
Mike Woodworth
mike at divergentmedia dot com
On Jan 31, 2006, at 9:40 PM, Lars Jensen wrote:
The idea of OCR'ing from an ISIGHT camera is pretty nifty,
but will be a LOT of work. The Susser already OCRs from puzzle
images, but it cheats and expects the puzzle to be perfectly aligned.
Then tell people to align the puzzle when they hold it up (perhaps
displaying an alignment line to help them). Or let them snap the
picture and click on the corners of the puzzle in the image, and then
transform suitably to input to your existing OCR.
First rev will satisfy perhaps 5% of those who try it. Use the revenue
from those 1,000,000 people to improve subsequent versions. When word
gets out that you don't have to do data entry, or rip pages out of
your book and feed them into a scanner, your usage will skyrocket. You
will share the stage at next year's Macworld show with Steve Jobs, and
can ride in on your diamond-encrusted Harley.
lj
_______________________________________________
Unsubscribe or switch delivery mode:
<http://www.realsoftware.com/support/listmanager/>
Search the archives of this list here:
<http://support.realsoftware.com/listarchives/lists.html>
_______________________________________________
Unsubscribe or switch delivery mode:
<http://www.realsoftware.com/support/listmanager/>
Search the archives of this list here:
<http://support.realsoftware.com/listarchives/lists.html>
|