NetNews Usenet Archive 1992 #20/NN_1992_20.iso/spool/comp/ai/neuraln/3451

Text File | 1992-09-10 | 16.7 KB | 421 lines

Newsgroups: comp.ai.neural-nets
Path: sparky!uunet!gatech!darwin.sura.net!jvnc.net!nuscc!iti.gov.sg!news
From: [e-mail] (Chan Kay Kit)
Subject: Information on OCR products: a summary
Message-ID: <[e-mail]>
Sender: [e-mail] (News Admin)
Reply-To: [e-mail]
Organization: Information Technology Institute, National Computer Board, S'pore
Date: Fri, 11 Sep 1992 06:22:02 GMT
Lines: 409


Hello everybody {:-)

    Thanks to all those who responded to my call for help on OCR
products. I have appended below all the information I received for anyone
who might be interested:

+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
From [e-mail] Wed Sep  2 18:53:27 1992


Please ask [e-mail] for information about the OCR capabilities of
PaperSight on the NeXT.  He can point you to customers who could supply the
kind of evaluation you are looking for.

Alan

+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
From [e-mail] Wed Sep  2 02:44:34 1992


Hi,

I read your query on comp.ai.neural-nets.

There is a company in Austria - Europe. They sell a very good product
called "proLector". Get in contact with them.

Sorry, I have only the posting address:

improx
Rennweg 83
A-2345 Brunn am Gebirge
Tel: [phone]
FAX: [phone]

regards,
Franz Peter SEILER

+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
From [e-mail] Tue Sep  1 18:39:02 1992


Hi!
you can find shareware software on OCR, in the SIMTEL20,
and you can download them. try to find the file OCRSHR22.ZIP.
and there is something more in the same directory.

there you can find a documantation file about OCR. in the OCRSHARE22 file.




            Bye
I will be glad if you can send me the mails you will get about OCR.


            my e-mail is : [e-mail]

+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
From [e-mail] Tue Sep  1 19:43:31 1992


The paper

"A Report on the Accuracy of OCR Devices"

by Stephen V. Rice, Junichi Kanai and Thomas A. Nartker
Information Science Research Institute
University of Nevada, Las Vegas
4505 Maryland Parkway
Las Vegas, NV 89154-4021

maybe useful.

It compares
Caere Omni Page Professional
Calera RS 9000
ExperVision TypeReader
Kurzweil 5200
Recognita Plus
and Toshiba ExpressReader

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
From @scapa.cs.ualberta.ca:[e-mail] Tue Sep  1 07:39:26 1992


Since you mentioned noise, you might be interested in our OCR demo
for adaptive logic networks (ALNs).  It will be a while before anything
commercial is available, but three people have said they want to produce
ALN-OCR systems.

Sorry if this is not what you are interested in.

Best wishes,

Bill


The atree Adaptive Logic Network Simulation Package and References

The latest release of the atree adaptive logic network (ALN)
simulation package, atree release 2.7, is available via anonymous ftp
from menaik.cs.ualberta.ca [129.128.4.241] in directory pub/ (use
binary mode).  It runs on IBM PCs and compatibles under Windows 3.x.
Included is documentation and ON-LINE HELP that will help you to
understand the basic principles of adaptive logic networks and enable
you to try out some examples of your own creation.  All C and C++
source code is provided.

The atree package is not a toy, despite the fact that it is used for
demonstration purposes, and is non-commercial.  Experimenters are
using it on challenging problems of medicine, physics and the
environment.  It has been used to grade beef based on ultrasound
images, design hardware to discriminate particles produced by an
accelerator, help to design walking prostheses for spinal cord damaged
patients and measure the composition of tarsands from spectral data.

It is possible to use inexpensive, off-the-shelf programmable logic
devices to realize the results of training ALNs in high-speed
hardware, though we do not provide those facilities in the atree
software.

Please read the license, and the warranty (that protects the
developers, not the users).  All neural networks which are "black
boxes" trained and tested on sample data are possibly unsafe, i.e.
unexpected outputs can occur. The current free atree package is no
exception.  A commercial version is planned for later release, and
eventually a safe design methodology will be available.

We shall use release 2.7 as an example here, but please see what the
latest version number is before you ftp the software by examining the
contents of pub/.  Atree release 2.7 is available in either of two
files in this directory: atre27.exe and a27exe.exe.  The file
atre27.exe contains the full C and C++ sources for those who want to
program some applications using Borland C++ 3.x and Application
Frameworks, or some similar C/C++ development environment.  The other,
smaller file contains just the executables for those who just want to
try out atree release 2.7.

Everyone should have a look at the OCR demo!  It has been referred to
as "quite impressive" by experts in the OCR area.  Test yourself
against the trained ALNs, and scribble in your own characters (any
language, any alphabet) to see how noisy and distorted the characters
can be, yet still be recognized by the logic networks.  The demo can
be obtained without the rest in pub/a27ocr.exe.

To set up your software on the PC under Windows 3.x, it is recommended
that you execute atre27.exe in your main directory, whereupon it will
create a subdirectory atree_27 and extract everything into it.
Running "setup" in that directory will create a group of icons you can
use to invoke demos and the facilities for programming adaptive logic
network applications in the lf language.  The "Open" command gives you
access to numerous instructive examples.  Clicking on the Help button
gives you access to explanations of theory and code.

The Unix version, atree release 2.0, will be upgraded sometime to add
some of the functionality of release 2.7.  It is not yet capable of
the nice color displays of the Windows version, but besides running on
Unix workstations it can be easily ported to Macintosh, Amiga, and
other machines.

Thanks to all those who have sent comments about their work on ALNs,
and have helped us to develop new concepts at a rapid rate. We would
appreciate your continued help in finding any problems in the latest
release.

Welcome to the world of adaptive logic networks!

Bill

***************************************************
Prof. William W. Armstrong, Computing Science Dept.
University of Alberta; Edmonton, Alberta, Canada T6G 2H1
[e-mail] Tel[phone] FAX [phone]
***************************************************

    Recommended Publications on Adaptive Logic Networks

W. Armstrong, Adaptive Boolean Logic Element, U. S. Patent 3934231,
Feb. 28, 1974 (multiple filings in various countries),
assigned to Dendronic Decisions Limited.

G. v. Bochmann, W. Armstrong, Properties of Boolean Functions
with a Tree Decomposition, BIT 13, 1974. pp. 1-13.

W. Armstrong, Gilles Godbout: Use of Boolean Tree Functions to Perform
High-Speed Pattern Classification and Related Tasks, Dept. d'IRO,
Universite de Montreal, Doc. de Travail #53, 1974. (unpublished,
except in summary form as follows:)

W. Armstrong and G. Godbout, "Properties of Binary Trees
of Flexible Elements Useful in Pattern Recognition", IEEE 1975
International Conf. on Cybernetics and Society, San Francisco, 1975,
IEEE Cat. No. 75 CHO 997-7 SMC, pp. 447-449.

W. Armstrong and J. Gecsei, "Architecture of a Tree-based
Image Processor", 12th Asilomar Conf. on Circuits, Systems
and Computers, Pacific Grove, Calif., 1978, pp. 345-349.

W. Armstrong and J. Gecsei, "Adaptation Algorithms for
Binary Tree Networks", IEEE Trans. on Systems, Man and
Cybernetics, 9, 1979, pp. 276-285.

W. Armstrong, J.-D. Liang, D. Lin, S. Reynolds, Experiments Using
Parsimonious Adaptive Logic, Tech. Rept. TR 90-30, Department of
Computing Science, University of Alberta, Edmonton, Alberta, Canada,
T6G 2H1.  This is now available in a revised form via anonymous FTP
from menaik.cs.ualberta.ca [129.128.4.241] in pub/atree2.ps.Z (the
title of the revised document is "Some Results concerning Adaptive
Logic Networks").

W. Armstrong, A. Dwelly, J.-D. Liang, D. Lin, S. Reynolds, Learning
and Generalization in Adaptive Logic Networks, in Artificial Neural
Networks, Proceedings of the 1991 International Conference on
Artificial Neural Networks ( ICANN'91), Espoo,
Finland, June 24-28, 1991, T. Kohonen, K.Makisara, O. Simula, J. Kangas eds.
Elsevier Science Publishing Co. Inc. N. Y. 1991, vol. 2, pp. 1173-1176.

Allen G. Supynuk, William W. Armstrong, Adaptive Logic Networks and
Robot Control, Proc. Vision Interface Conference '92, also called
AI/VI/GI '92, Vancouver B. C., May 11-15, 1992, pp. 181 - 186.

R. B. Stein, A. Kostov, M. Belanger, W. W. Armstrong and D. B.
Popovic, Methods to Control Functional Electrical Stimulation in
Walking, First International FES Symposium, Sendai, Japan, July 23 -
25, 1992 (Invited paper, to appear).

Aleksandar Kostov, Richard B. Stein, William W. Armstrong, Monroe
Thomas, Evaluation of Adaptive Logic Networks for Control of Walking
in Paralyzed Patients, 14th Ann. Int'l Conf. IEEE Engineering in
Medicine and Biology Society, Paris, France, Oct. 29 - Nov. 1, 1992
(to appear).

In case you have difficulty in obtaining the above documents or the
atree release 2.0 software for Unix or release 2.7 for IBM-PC and
compatibles under Windows, the software and all of the above documents
prior to 1991 can be obtained from the University of Alberta for a
media fee to cover the costs of copying and mailing of $150
(Canadian), made payable to the University of Alberta.  Two 3 1/2"
diskettes are normally included but an attempt will be made to satisfy
needs for other media, e.g. tapes.  Orders can be sent c/o Professor
W. W.  Armstrong, Department of Computing Science, University of
Alberta, Edmonton, Alberta, Canada, T6G 2H1.

+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
From [e-mail] Tue Sep  1 07:38:26 1992


Some time ago we went through n exercise of looking at OCR products for
the PC.  We wanted a scanner and OCR software.  We eventually purchased
Wordscan from Calera Recognition Systems, 2500 Augustine Drive, Santa
Clara, CA, USA, 95054.  This runs under Windows and works very well.

The scanner we bought is a Hewlett Packard ScanJet IIc, a flat bed A4.
Wordscan will interpret a well printed and formatted sheet with 100%
accuracy.  It will also put it out in any one of 20 or 30 word processor
formats.  It retains bolding, underlining, italics, point sizes etc, etc.
You can also switch that off if you wish.

I cannot speak for others such as OmniPage (the only other one in the
race that I could find) as when I tried to run the demo we got with the
scanner it would not work with windows running in standard or real mode,
and, at the time, I could not run windows in enhanced mode.

I would be interested in the features you discover about other packages.
Are you planning to produce a report or list of your findings?
If so I would like to receive a copy.

If I can give you any more information please let me know.

Best regards  Garth Tier

--
Garth Tier who is in Alice Springs not Canberra!

+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
From [e-mail] Tue Sep  1 00:33:16 1992


  What platform are you going to be working on? Unix or PC or Macintosh or ??
  For the PC, PC Magazine and other PC publications have published reviews
  on OCR systems that go into a good amount of detail on their capabilities.
  I have not seen a review for a Unix based OCR yet.
   Many of the points you addressed before were given in the PC articles.
   If you are running on a PC platform, I can dig out a my old PC mags and
   let you know the volume number for the OCR review article.

  The reason I am (somewhat) knowledgable on the subject is that at Cal Poly
 some students (myself included) just finished a undergraduate class in
 computer vision applications. We wrote a full optical character recogntion
 system that can take in a text image, checks to see if their is any rotation
 on the image (if the text was scanned in say at a 5 degree angle) and
 rotates the image back to the proper upright position. This step has
 increased the average accuracy of recognition rates. A few weeks ago I posted
 my own software up for public ftp on polyslo.csc.calpoly.edu.
  At the end of the class I was obtaining greater than 99% accuracy on
 non-rotated images, and 98% acc. on rotated images. My program does
 auto-binarization, so that a dark background, or lots of static will
 be filtered out (such as in low quality fax transmission). Some other
 students went even further that myself and were able to obtain 98%+
 recognition rates even on very hazy/noisy images. I only recieved a B
 in the class, so you can see that my project was not top of the line.
 Other students were regularly obtaining full, 100% acc. on non-rotated
 images, and 99+% on rotated images. Currently the project only handles one
 font type at a time (the neural net COULD be trained on multiple fonts, but
 this could bring the recog rates down to around 97%) but since size is not
 one of the paramters used for the net, different point size fonts can be
 used with reasonably the same rates of recognition.
   If you would like more information then what I have provided, feel free
 to mail me, or you can also send mail to the professor who taught the class,
 as he would most likely have many sources of reference for this type of
 material. His email address is:
    [e-mail]

 Good Luck.

Shawn Jacques
[e-mail]
[e-mail]

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
From [e-mail] Mon Aug 31 15:17:36 1992


Hi :)
>
>     There are many commercial OCR engines on the market right now for
> reading printed text. I am aware of products like the Kurzweil K-5000 and
> Discover 7320 Model 5 (from Xerox), OmniPage (from Caere Corp.), SAIC (from
> Differential Processing System) and NestorReader.
>
Those are on Mac and PC. On other computers like Amiga, there is some also.
I must tell that I have used only OmniPage, and heard of some others...

>     I wonder if anybody has done independent evaluation of any OCR
> products, since such information would be useful to potential users.
> Specifically, I need answers to the following questions:
>
> (i) name of product and company which produces it

Not quite sure about the name, but it's made by Miggraph. They sell it with
their own scanner, but the product works with files too.
>
> (ii) fonts and point sizes handled by product
>
>From 6 up to 127 if I remember right. Maybe even bigger.

> (iii) whether the OCR engine exhibits true multifont and "multisize"
>       capability i.e. is able to recognize text without the user
>       having to specify in advance its font type and point size.

Yes, I think so, as I remember reading that it handles multiple fonts, and
has some learning algorithm.
>
>       If so, what is the accuracy over a variety of fonts and point sizes?
>
>       If not, what is the accuracy over common fonts like Courier, Times
>       Roman and Helvetica and point sizes from 8 to 24 points?
>
I don't know about those numbers. All fonts are included in the starter
package.

> (iv) What are the commercial OCR products, if any, that can
>      1. recognise a page of text containing a mixture of fonts and point
>     sizes at greater than 98% accuracy?
>
>      2. recognise a page of text containing a single known font and a
>     range of point sizes at greater than 99% accuracy?
>
No comments.

> (v) Are there any commercial OCR systems that accept noisy input or
>     text on shaded background eg. carbon and photostated/fax copies and
>     still maintain an acceptable accuracy (say greater than 95%)?

Not sure, but I think that software can find text in IFF files. Not quite sure.
And it can read Fax'd files too.

> Further, if there are any evaluations of commercial handwritten OCR systems,
> I will also be interested to hear about them.
>
> Please email all replies/references to [e-mail]. If there are
> enough responses, I will summarise.

Please do. And I have no other knowledge of that Miggraph product except
what I have read from some news.
>
> Thanks!
>
YRWLCM

Esa Haapaniemi
University of Oulu
Finland

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++


Kay Kit



---
Kay-Kit CHAN                | Internet: [e-mail]
Knowledge Systems Lab            | Bitnet:   kaykit@itivax
Information Technology Institute    | Tel:      [phone]
National Computer Board of Singapore    | Fax:      [phone]