John Riehl realman@smart.net
---------- Forwarded message ----------
Date: Tue, 29 Apr 1997 09:50:28 -0400
From: Deane Blazie <deane@blazie.com>
Reply-To: optacon@blazie.com
To: Multiple recipients of list <optacon@blazie.com>
Subject: Grant Progress
Hi Optacon listers.
This is an update on what's happening on the Grant and a request for some
input from you and some help.
We have software now that will allow us to scan a document into memory and
to display the image with and without various image transformations. We
can view and print the origonal image. We can view and print the image
after the characters in the image have been separated more than the
scanned image. You would expect this to be easier to read on some
documents where some letters are very close together. We can view and
print images in a different font (font substitution). With this one we
can reduce all characters to the same size and into a font that is much
more readable (we think) than the scanned font. Of course this is very
dependant on the font of the origonal document. Another transform allows
us the take the scanned image and view/print it after the characters have
been whittled down to only thin lines. We call this skeletonization
(sp?). This one needs some work I think because it also makes the
characters more ragged looking.
We are almost ready to begin testing the change in reading rates after
these transformations have been applied to a scanned image.
Now for some help that I need. First of all I need some sample documents
that are particularly difficult to read. The problem is that if we take a
normal "type written" document, the image processing would show little or
not improvements because a "type written" document is pretty easy to read
already and you wouldn't expect much improvement. So, we need some real
life documents that are difficult to read or at least troublesome to read.
Any help in this area would be appreciated. You can send documents to:
Deane Blazie
Blazie Research, Inc.
2850 Market Place
Stuart, Fl 34997
While I can't promise, I will try to scan your document (one page) and
send you a copy of the various transformations in print. YOu can then
comment on the "VALUE" of the transformations.
By the way, we haven't gone off the deep end on this. We fully intend to
have the primary operating mode of a new device be the normal
"unprocessed" image you are all familar with. You may be able to improve
a document by turning on some of this fancy image processing. IF YOU WANT
TO.
We are also making progress on a new way to make the tactile array. I
don't have much in the way of details on this but some of our experiments
have been VERY encouraging. The main idea with this research is to make
it easier (less expensive) to make the array. We are also experimenting
with a larger array. More later.
So, that's the update. ANy help would be appreciated.
Deane Blazie
Jim Rodgers
This archive was generated by hypermail 2b29 : Sun Dec 02 2012 - 01:30:04 PST