kopia lustrzana https://gitlab.com/sane-project/website
73 wiersze
3.3 KiB
HTML
73 wiersze
3.3 KiB
HTML
<!-- received="Wed Nov 12 17:44:29 1997 PST" -->
|
|
<!-- sent="Thu, 13 Nov 1997 09:46:46" -->
|
|
<!-- name="Colin 't Hart" -->
|
|
<!-- email="hart@cygnus.uwa.edu.au" -->
|
|
<!-- subject="Re: OCR Software..?!" -->
|
|
<!-- id="3.0.3.32.19971113094646.00a91ab0@cygnus.uwa.edu.au" -->
|
|
<!-- inreplyto="199711130017.TAA14813@lemur.magnet.com" -->
|
|
<title>sane-devel: Re: OCR Software..?!</title>
|
|
<h1>Re: OCR Software..?!</h1>
|
|
<b>Colin 't Hart</b> (<a href="mailto:hart@cygnus.uwa.edu.au"><i>hart@cygnus.uwa.edu.au</i></a>)<br>
|
|
<i>Thu, 13 Nov 1997 09:46:46</i>
|
|
<p>
|
|
<ul>
|
|
<li> <b>Messages sorted by:</b> <a href="date.html#101">[ date ]</a><a href="index.html#101">[ thread ]</a><a href="subject.html#101">[ subject ]</a><a href="author.html#101">[ author ]</a>
|
|
<!-- next="start" -->
|
|
<li> <b>Next message:</b> <a href="0102.html">Alex Belits: "Re: OCR Software..?!"</a>
|
|
<li> <b>Previous message:</b> <a href="0100.html">Andrew Kuchling: "Re: OCR Software..?!"</a>
|
|
<li> <b>In reply to:</b> <a href="0100.html">Andrew Kuchling: "Re: OCR Software..?!"</a>
|
|
<!-- nextthread="start" -->
|
|
<li> <b>Next in thread:</b> <a href="0102.html">Alex Belits: "Re: OCR Software..?!"</a>
|
|
<li> <b>Reply:</b> <a href="0102.html">Alex Belits: "Re: OCR Software..?!"</a>
|
|
<!-- reply="end" -->
|
|
</ul>
|
|
<!-- body="start" -->
|
|
At 19:17 12/11/97 -0500, Andrew Kuchling wrote:<br>
|
|
<p>
|
|
<i>> Then the image</i><br>
|
|
<i>>has to be segmented into words, and words into letters; each letter is</i><br>
|
|
<i>>then recognized, and usually a confidence value is attached to each</i><br>
|
|
<i>>letter.</i><br>
|
|
<p>
|
|
Apple was doing work on handwriting recognition and had phenomenally<br>
|
|
high recognition rates by looking at words rather than letters. Of<br>
|
|
course this is not language independent and requires dictionaries -<br>
|
|
but these can be built up by the software (training).<br>
|
|
<p>
|
|
An ideal OCR would use the "lower level" results from the letter<br>
|
|
recognition to influenece the "high level" word recognition and<br>
|
|
vice versa.<br>
|
|
<p>
|
|
I'm not familiar with this Neural Net speak, but I think this<br>
|
|
sort of thing is possible and should help to achieve much higher<br>
|
|
accuracy levels.<br>
|
|
<p>
|
|
Cheers,<br>
|
|
<p>
|
|
Colin<br>
|
|
<p>
|
|
---------------------------------------------------------------<br>
|
|
Colin 't Hart, Systems Analyst Email: <a href="mailto:cthart@gbs.com.au">cthart@gbs.com.au</a><br>
|
|
Geographic Business Systems Pty Ltd<br>
|
|
The Garden Office Park Phone: +61 8 9443 8822<br>
|
|
345 Harborne St (PO Box 761) Fax: +61 8 9242 4412<br>
|
|
Herdsman WA 6017 <br>
|
|
<p>
|
|
<pre>
|
|
--
|
|
Source code, list archive, and docs: <a href="http://www.mostang.com/sane/">http://www.mostang.com/sane/</a>
|
|
To unsubscribe: echo unsubscribe sane-devel | mail <a href="mailto:majordomo@mostang.com">majordomo@mostang.com</a>
|
|
</pre>
|
|
<!-- body="end" -->
|
|
<p>
|
|
<ul>
|
|
<!-- next="start" -->
|
|
<li> <b>Next message:</b> <a href="0102.html">Alex Belits: "Re: OCR Software..?!"</a>
|
|
<li> <b>Previous message:</b> <a href="0100.html">Andrew Kuchling: "Re: OCR Software..?!"</a>
|
|
<li> <b>In reply to:</b> <a href="0100.html">Andrew Kuchling: "Re: OCR Software..?!"</a>
|
|
<!-- nextthread="start" -->
|
|
<li> <b>Next in thread:</b> <a href="0102.html">Alex Belits: "Re: OCR Software..?!"</a>
|
|
<li> <b>Reply:</b> <a href="0102.html">Alex Belits: "Re: OCR Software..?!"</a>
|
|
<!-- reply="end" -->
|
|
</ul>
|