Clara OCR 0.9.9
---------------

Clara OCR version 0.9.9 is available. Clara OCR is designed to
OCR old books. It's available under the terms of the GPL.

Clara OCR features a GUI and a web interface. It also features a
command-line interface, but X Windows is always required both to
compile and run. The development effort is using Intel 32-bits
boxes running various Linux distributions and FreeBSD.

Clara OCR can be found at http://www.claraocr.org

New in 0.9.9
------------

1. Added grayscale native support (PGM format), plus four
binarization methods.

2. Added internal preprocessor including deskewing, balancing,
thresholding and interpolation (by Giulio Lunati).

3. Packaged for NetBSD (thanks Thomas Klausner).

4. Various new features: border path computing, barcode search,
detection of extremities, PAGE only mode, the flea, the spyhole,
instant threshold, per-depth optimized X code, etc.

5. Many bugfixes and interface enhancements.

6. Documentation updated (but not finished). Glossary added.

Obs:

* The CHANGELOG is available at http://www.claraocr.org/CHANGELOG

* The FAQ is available at http://www.claraocr.org/FAQ.html

* The archives of the list claraocr-devel can be acessed through
the SmartList native features (visit
http://www.claraocr.org/lists.html for details).

* Once a week the development files are copied to the net and
made available through FTP. News are posted to the lists
claraocr-devel and claraocr-announce on a weekly basis.


-------------------------------------------------------------

ANNOUNCE of version 0.9.8:

Clara OCR 0.9.8
---------------

Clara OCR version 0.9.8 is available. Clara OCR is designed to
OCR old books. It's available under the terms of the GPL.

Clara OCR features a GUI and a web interface. It also features a
command-line interface, but X Windows is always required both to
compile and run. The development effort is using Intel 32-bits
boxes running various Linux distributions and *BSD.

Clara OCR can be found at http://www.claraocr.org

New in 0.9.8
------------

1. Thresholding facility added

Now the binarization of grayscale images can be automatically
tuned by the selthresh.pl script. Read the section "A first OCR
project" of the Advanced User's Guide for details.

2. New classifier and new skeleton-computing heuristics.

A third classifier (called "pixel distance") was created. Two new
skeleton-computing heuristics (#5 and #6) were added.

3. The Advanced User's Guide (partially finished) was carefully
updated.

4. FreeBSD supported.

5. Various small enhancements. In particular, the features to
organize the various fonts used along one same book (see the
CHANGELOG).

6. Various bugfixes.

7. Skeleton code is about 40%-50% faster.

Obs:

* The CHANGELOG is available at http://www.claraocr.org/CHANGELOG

* The FAQ is available at http://www.claraocr.org/FAQ.html

* The archives of the list claraocr-devel archives can be acessed
through the SmartList native features (visit
http://www.claraocr.org/lists.html for details).

* Once a week the development files are copied to the net and
made available through FTP. News are posted to the lists
claraocr-devel and claraocr-announce on a weekly basis.

-------------------------------------------------------------

ANNOUNCE of version 0.9.7:

Clara OCR - powerful free/open-source OCR program
-------------------------------------------------

The Clara OCR Project is proud to announce the availability of
its new version 0.9.7, associated with a new web site and mailing
lists.

Clara OCR is an Optical Character Recognition (OCR) program. It
features both a powerful GUI for the X Window System, and a web
interface. The web interface is able to collect revision efforts
from the Internet, using a simple revision model.

Clara OCR is intended to be used in the cooperative optical
recognition of old books. It tries to facilitate fine-tuning, so
an optical recognition project is enabled to invest resources in
tuning the OCR, in order to achieve better recognition results
for one specific book, and reduce the overall revision cost.

Clara OCR is distributed under the terms of the GNU General
Public License.

The development started in the second half of 1999. After two
years of intensive work, it is currently close to the 1.0
version. In some cases, the current beta version could actually
be used for production.

The documentation is almost complete. We already have a FAQ, a
Tutorial, and preliminary versions of the Advanced User's Manual
and the Developer's Guide.

The development effort is using Intel 32-bit boxes and various
different distributions of GNU/Linux.

Clara OCR can be found at http://www.claraocr.org.