kopia lustrzana https://gitlab.com/sane-project/website
93 wiersze
4.7 KiB
HTML
93 wiersze
4.7 KiB
HTML
|
<!-- received="Sat Jun 21 19:15:33 1997 MST" -->
|
||
|
<!-- sent="Sun, 22 Jun 1997 01:32:35 +0200 (MET DST)" -->
|
||
|
<!-- name="becka@sunserver1.rz.uni-duesseldorf.de" -->
|
||
|
<!-- email="becka@sunserver1.rz.uni-duesseldorf.de" -->
|
||
|
<!-- subject="Re: Problem compiling umax-driver with glibc-2.03" -->
|
||
|
<!-- id="m0wfZdz-000BWNC@charon.beck-sw.de" -->
|
||
|
<!-- inreplyto="199706212009.QAA05718@lemur.magnet.com" -->
|
||
|
<title>sane-devel: Re: Problem compiling umax-driver with glibc-2.03</title>
|
||
|
<h1>Re: Problem compiling umax-driver with glibc-2.03</h1>
|
||
|
<a href="mailto:becka@sunserver1.rz.uni-duesseldorf.de"><i>becka@sunserver1.rz.uni-duesseldorf.de</i></a><br>
|
||
|
<i>Sun, 22 Jun 1997 01:32:35 +0200 (MET DST)</i>
|
||
|
<p>
|
||
|
<ul>
|
||
|
<li> <b>Messages sorted by:</b> <a href="date.html#131">[ date ]</a><a href="index.html#131">[ thread ]</a><a href="subject.html#131">[ subject ]</a><a href="author.html#131">[ author ]</a>
|
||
|
<!-- next="start" -->
|
||
|
<li> <b>Next message:</b> <a href="0132.html">Guenther Beulshausen: "Problems with sane-0.6/UMAX S-12 and gimp 0.99.10"</a>
|
||
|
<li> <b>Previous message:</b> <a href="0130.html">orauch@physik.uni-osnabrueck.de: "umax-backend"</a>
|
||
|
<li> <b>In reply to:</b> <a href="0128.html">Andrew Kuchling: "Re: Problem compiling umax-driver with glibc-2.03"</a>
|
||
|
<!-- nextthread="start" -->
|
||
|
<li> <b>Next in thread:</b> <a href="0129.html">becka@sunserver1.rz.uni-duesseldorf.de: "Re: Problem compiling umax-driver with glibc-2.03"</a>
|
||
|
<!-- reply="end" -->
|
||
|
</ul>
|
||
|
<!-- body="start" -->
|
||
|
Hi !<br>
|
||
|
<p>
|
||
|
<i>> My summer project is to write some sort of OCR software so I</i><br>
|
||
|
<i>> can help the Gutenberg Project; running Linux/AXP only, I'm not about</i><br>
|
||
|
<i>> to buy NT just for this one thing, and it's a good chance to learn</i><br>
|
||
|
<i>> image processing and neural networks. Image processing textbooks and</i><br>
|
||
|
<i>> C.M. Bishop's _Neural Networks For Pattern Recognition_ are in hand,</i><br>
|
||
|
<i>> so all I need is to buy a scanner that works with SANE and start</i><br>
|
||
|
<i>> working on code.</i><br>
|
||
|
<p>
|
||
|
Yeah - would like to see that, too.<br>
|
||
|
<p>
|
||
|
<i>> 3) Scale each letter down to a 32x32 array of pixels. Use</i><br>
|
||
|
<i>> Kohonen-Loeve feature extraction to derive a 32-element vector from</i><br>
|
||
|
<i>> that array; it's faster to run a neural network over a 32-element</i><br>
|
||
|
<i>> vector than 1024 pixel values. </i><br>
|
||
|
<p>
|
||
|
I don't have references on this feature-extractor handy, but from the<br>
|
||
|
people involved it should be good. <br>
|
||
|
<p>
|
||
|
Pay close attention on which feature-extractor(s) to use. This is the <br>
|
||
|
most important thing with pattern recognition.<br>
|
||
|
<p>
|
||
|
<i>> 4) Use a neural network to classify the vector, getting a</i><br>
|
||
|
<i>> letter and a confidence value.</i><br>
|
||
|
<p>
|
||
|
A really well chosen feature-extractor should allow the use of simpler<br>
|
||
|
methods like vector-matching.<br>
|
||
|
<p>
|
||
|
The general guideline is : Make up a good feature-extractor.<br>
|
||
|
While neural networks are a nice concept, you often cannot know,<br>
|
||
|
if they train well for a given set of patterns.<br>
|
||
|
<p>
|
||
|
I have seen a lot of software which relied too much on the NN and thus<br>
|
||
|
didn't work very well for low-quality data as often got from scanners.<br>
|
||
|
<p>
|
||
|
If you encounter problems with low contrast or noisy images, I could<br>
|
||
|
contact one of our profs which is into image-enhancement.<br>
|
||
|
<p>
|
||
|
<i>> What could SANE do to support OCR applications in particular?</i><br>
|
||
|
<i>> >From my reading, I can't really think of anything special that the</i><br>
|
||
|
<i>> backend or frontend should support, but that may just indicate that I</i><br>
|
||
|
<i>> haven't done enough research.</i><br>
|
||
|
<p>
|
||
|
The most important thing is to be able to control the scanner <br>
|
||
|
(especially sheet-feeders etc.) from the ocr app. SANE has support for this.<br>
|
||
|
<p>
|
||
|
CU,Andy<br>
|
||
|
<p>
|
||
|
<pre>
|
||
|
--
|
||
|
Andreas Beck | Email : <<a href="mailto:becka@sunserver1.rz.uni-duesseldorf.de">becka@sunserver1.rz.uni-duesseldorf.de</a>>
|
||
|
<p>
|
||
|
<pre>
|
||
|
--
|
||
|
Source code, list archive, and docs: <a href="http://www.azstarnet.com/~axplinux/sane/">http://www.azstarnet.com/~axplinux/sane/</a>
|
||
|
To unsubscribe: mail -s unsubscribe <a href="mailto:sane-devel-request@listserv.azstarnet.com">sane-devel-request@listserv.azstarnet.com</a>
|
||
|
</pre>
|
||
|
<!-- body="end" -->
|
||
|
<p>
|
||
|
<ul>
|
||
|
<!-- next="start" -->
|
||
|
<li> <b>Next message:</b> <a href="0132.html">Guenther Beulshausen: "Problems with sane-0.6/UMAX S-12 and gimp 0.99.10"</a>
|
||
|
<li> <b>Previous message:</b> <a href="0130.html">orauch@physik.uni-osnabrueck.de: "umax-backend"</a>
|
||
|
<li> <b>In reply to:</b> <a href="0128.html">Andrew Kuchling: "Re: Problem compiling umax-driver with glibc-2.03"</a>
|
||
|
<!-- nextthread="start" -->
|
||
|
<li> <b>Next in thread:</b> <a href="0129.html">becka@sunserver1.rz.uni-duesseldorf.de: "Re: Problem compiling umax-driver with glibc-2.03"</a>
|
||
|
<!-- reply="end" -->
|
||
|
</ul>
|