ABBYY FINEREADER ENGINE 10 TECHNICAL RELEASE NOTES

burgerraraSoftware and s/w Development

Nov 18, 2013 (3 years and 11 months ago)

5,382 views

ABBYY FINEREADER ENG
INE 10

TECHNICAL RELEASE NO
TES

R1


PART NUMBER, BUILD N
UMBER

Part#


681/15

Build#


10.0.3
.
494

WHAT’S

NEW

EXTREME RECOGNITION
SPEED


Feature



Description



Benefit


Tuned Fast Mode

Adjusted Fast mode achieves approximately 100%
speed increase* with more than 98.5% accuracy
for most European languages.

Developers obtain a unique
speed
-
accuracy rate for popular
European languages.

*compared to Fast mode o
f ABBYY FineReader Engine 9.0 (First release, 21 October 2008)


Note:

ABBYY's multi
-
core support architecture ensures close to linear performance growth with increasing
number of cores for multipage documents.

Improved Recognition Accuracy


Feature



Description



Benefit


Improved classifier for
CJK

The recognition accuracy for Chinese, Japanese
and Korean languages increased 30%
-
50% due to
an improved Asian character classifier.

Developers can use a high
-
quality Asian OCR together with
dozens other languages (198 in
all) in one SDK.

New mode for low
resolution scans

The new recognition mode for low quality
documents
-

old faxes, low resolution scans
provides 20% higher accuracy for such documents
in comparison to standard Normal mode in

previous versions.

Significantly increase recognition
for low resolution images.

*comparing to Normal mode of ABBYY FineReader Engine 9.0 (First release, 21 October 2008)

IMPROVED API

Feature

Description

Benefit

Special profiles for
popular recognition
scenarios

New profiles are provided in this release. These
new profiles are predefined with optimal
parameters for faster testing and development.
Predefined profiles can be used out
-
of
-
the
-
box, or
parameters and tuning can be changed manually.

Developers
who do not want to
spend the time studying the vast
array of ABBYY SDK API's
features and parameters can
choose a profile and rapidly
integrate ABBYY OCR
technologies into your
application.

Document structure API

ABBYY FineReader Engine 10 provides a uniq
ue
feature
-
set for access to document structure
elements like headings, chapters, page numbers,
footnotes, headers, footers and so on.

New API functions allow precise
and careful content editing,
saving the structure of original
document.


UNIQUE LAYOUT
RECONSTRUCTION ABILI
TIES

Feature

Description

Benefit

Document

structure

detection

ABBYY FineReader Engine 10 automatically detects
headings in recognized documents, determines their
level in the document structure, defines their text styles
and
reconstructs them as a Document Map of the
resulting document.

Developers can carefully
reconstruct the original layout of
the document and its structure
for easy content reuse

TOC

reconstruction

This version delivers final documents where the Table of
Co
ntents exists as a set of links to the headings. After
document editing, the TOC could be updated
automatically as a single object to add new headings
and revise page numbers.

Charts and

diagrams

detection

The automatic charts and diagrams detection
feature
was improved in this version. Now it is possible to
extract text on charts or retain the chart as an image on
the resulting document.


Picture and table
captions processing

ABBYY FineReader Engine 10 automatically detects
captions associated with
pictures and tables. The results
can be exported to the final document as a single frame
including the picture, caption and its title.


Defining

document

styles

The new version analyzes text font type, size, and its
placement and detects the corresponding

font style for
every type of text. So for the headings of each level
there are special styles, for ordinary text, for TOC and
for picture captions there are also special styles.


“Glossy magazine”
processing model

New ABBYY SDK can reconstruct complicate
d layouts
consisted of many pictures and text blocks on a page or
including very large pictures for the whole page



IMPROVED PDF EXPORT


Feature

Description

Benefit

Superior quality
-
size ratio
for PDF files

New PDF export together with improved MRC
(Mixed Raster Content) compression technology
achieves higher quality and reduced size of
converted PDF documents.

Higher quality PDFs and more
efficient utilization of storage.

PDF export profiles

There are more than 40 parameters for PDF export
tuning.
ABBYY FineReader Engine 10 provides
predefined profiles with optimal values for popular
export variants:

MaxQuality

Balanced

MinSize

MaxSpeed

With predefined PDF export
profiles developers
automatically set optimal values
for particular tasks.

UNRIVALED
DOCUMENT IMAGE AND P
HOTO PROCESSING

Feature

Description

Benefit

New features of Camera
OCR™

Camera OCR technology
-

the set of document
photo adjustment features for better recognition
results was improved with new features:

Automatic correction of 3D per
spective distortions

Blurred image correction

ISO noise reduction

The majority of document
images taken with common
camera optics have some
defects. With the new Camera
OCR features in ABBYY
FineReader Engine 10,
developers have access to image
enhancement

technologies
designed for better recognition
results for these specific issues.

New binarization

Previous ABBYY SDK versions provided very high
quality binarization, but in some of the most
difficult cases it could produce errors and loss of
information.

New binarization technologies ensure
complete text retention and prevent information
loss even in difficult cases.

Provides outstanding quality and
guarantees retention of valuable
image information, resulting in
higher OCR accuracy.

Color marks and
stamps
filtering

Document stamps or marks usually interfere with
text and decrease the quality of OCR results. New
in this version are special features for color marks
and stamps filtering which result in improved
recognition accuracy.

This is a useful fea
ture for data
capture systems, preventing
data loss from fields covered by
stamps and color marks.


FLEXIBLE

AND ROBUST PROTECTIO
N SYSTEM

Feature

Description

Benefit

Improved protection

The new protection system in ABBYY FineReader
Engine 10 provides
developers the ability to:

Delegate and control SDK usage rights in a local
network

Count and control the number of recognized
characters, pages, usage time and computing
power

Track and control SDK usage on terminal servers
and virtual machines

With the
unique abilities of the
new protection system,
developers can choose an optimal
licensing scheme so you pay only
for the features you use.

Developers can easily construct
their own protection system
based on ABBYY's integrated SDK
protection features, allo
wing total
control of your application usage.
Also, developers can measure
integrated OCR performance by
multiple units: pages per month
or per year, characters per
second, CPUs and workstations,
and so on. This guarantees
maximum profit from your
applicat
ion.


USEFUL AND CLEAR DEV
ELOPER'S DOCUMENTATI
ON

Feature

Description

Benefit

Improved Developers
Guide (Help)

Improved documentation includes an updated
structure and appearance, as well as new content
including general product descriptions, API
specifications, usage samples and best practices.

More inclusive and useful search
results within our documentation.


FINEREADER ENGINE 10
.0: UPGRADE FROM PRE
VIOUS VERSION

ABBYY FineReader Engine 10 works with any previous ABBYY FineReader Engine major ve
rsion installed on the
same PC if products' installation folders are different.

ABBYY FineReader Engine 10 requires a valid serial number for functioning.

ABBYY FineReader Engine 10 has certain incompatibilities with API implemented in the previous
versions described
in "ABBYY FineReader Engine 10 and 9.0 compatibility" section of the product Help file. Every existing customer
willing to upgrade his copy of ABBYY FineReader Engine should read the article first.

COMPATIBILITY VERSIO
N 9.0: NEW DEFAULT
INSTALLATION FOLDERS


Compared to the previous version, the release has different installation folders.

Folder

Description

"%ProgramFiles%
\
ABBYY SDK
\
10
\
FineReader Engine"

Default value for a folder storing binaries, the Help
file and the guides, and USB
dongle driver
redistribution. Also, the Read Me file is located there.
The user can change the destination during
installation process.

"%ProgramData%
\
ABBYY
\
SDK
\
10
\
Licenses"

This is the permanent place for protection subsystem
files.

"%ProgramData
%
\
ABBYY
\
SDK
\
10
\
FineReader Engine"

This is the permanent place for auxiliary Engine files
that include files and samples.


LICENSED 3
RD
-
PARTY SOFTWARE

This version uses several licensed 3
rd
-
party libraries. They enable the product with useful functionality

and require
us to add certain acknowledgements and items in the product documentation and/or License Agreement.

The list of newly licensed technologies is below.

FREETYPE FONT PARSIN
G LIBRARY

This version uses FreeType Open Source library for opening font

files. That gives us ability to be independent from
Windows system libraries and get more reliable results in PDF/A export and same time obliges us to specify certain
copyrights in the product documentation.

INTEL® INTEGRATED PE
RFORMANCE PRIMITIVES

This
version uses Intel® Integrated Performance Primitive (Intel® IPP) library for gaining product performance in
digital camera distortion correction (e.g. blur) and some document analysis calculations. That obliges us to specify
certain copyrights in the prod
uct documentation.

DISTRIBUTION COMPONE
NTS

DOCUMENTATION

Material

Language

File name

Description

Readme

English

CD
\
ABBYY SDK
\
10
\
FineReader Engine
\
Readme.htm

A short summary of the
distribution package and
the setup procedure.

Product

Installation

Guide

English

CD
\
ABBYY SDK
\
10
\
FineReader
Engine
\
Help
\
FREngine10AdminGuide.pdf

The guide describes how
to install the software
library.

License Server
Installation Guide

English

CD
\
License Server
\
ABBYY SDK 10 License
Server
\
LicenseServer10AdminGuide.pdf

The
guide describes how
to install the License
Server.

Help File

English

CD
\
ABBYY SDK
\
10
\
FineReader
Engine
\
Help
\
FREngine10.chm

A full and detailed
description of the
product

functionality. It also
includes chapters on
License Manager.

License Server
Help
File

English

CD
\
License Server
\
ABBYY SDK 10 License
Server
\
LicenseManager10.chm

A full and detailed
description of the
License Service and
License Manager.

User's Guide

English

CD
\

ABBYY SDK
\
10
\
FineReader
Engine
\
Help
\
FREngine10UserGuide.pdf

Printing
version of the
Help File.

Distribution List


English

CD
\
ABBYY SDK
\
10
\
FineReader
Engine
\
Help
\
FREngine10_Distribution.csv

A list of files to distribute
with description of
responsibility and
requirements to
distribute.

SAMPLE IMAGES

There are prepared
multi
-
page sample images in several languages:

1.

Chinese Simplified

2.

Dutch;

3.

English;

4.

French
-

5.

German;

6.

Italian;

7.

Polish;

8.

Russian;

9.

Swedish.


The distribution CD contains all images in the following folder:

CD
\
CommonAppData
\
ABBYY
\
SDK
\
10
\
FineReader Engine
\
Samples
\
SampleImages

CODE SAMPLES


The distribution contains samples described below in the following folder:

CD
\
CommonAppData
\
ABBYY
\
SDK
\
10
\
FineReader Engine
\
Samples

To view samples description and run them or open source fi
les a customer should use "Code Samples Library", the
HTML
-
based browsing utility, or go directly to the location of samples. A developer can find shortcuts for both the
utility and the samples folder under the Start menu after the installation process is
completed.


Hello

Available in
:

Delphi 5,

VB .Net,

VB 6,

C++ (COM)

C++,

C#.


This sample performs document conversion with just a few lines of code. This sample will help you to start
development using ABBYY SDK.


CustomLanguage

Available in
:

VB .Net,

VB
6,

C++ (COM)

C++,

C#.


This sample creates a new recognition language and changes its dictionary to a manually
-

created sample
dictionary. After recognition, calculates the number of words in the text and how many of them were found in the
user dictionary.

You can create a custom recognition language which will help your program to read specific types
of data correctly.


EventsHandling

Available in
:

VB .Net,

VB 6,

C++ (COM)

C++,

C#.


This sample illustrates the use of the callback interfaces using the FRDoc
ument callback interface
(IFRDocumentEvents) as an example. The sample shows the progress of recognition and export during image
processing. You can use the callback interfaces to control image processing.


RecognizedTextProcessing


Available in
:

VB .Net,

VB 6,

C++ (COM)

C++,

C#.


This sample calculates recognition statistics (e.g. the number of suspicious characters and rejects, the number of
words which are not in the dictionary). You can use information about uncertainly recognized characters and
words
for checking the results of recognition.


FREComWrapper

Available in
:

C++ (COM).


This sample provides an easy way to use ABBYY FineReader Engine from script languages. The sample produces
FREngineWrap.dll, which can be used to get an ABBYY FineReader Engi
ne object from a script language such as
VBScript, JavaScript, etc.


CommandLineInterface

Available in
:

C++.


This sample provides the command line interface of ABBYY FineReader Engine. The sample produces a
CommandLineInterface utility, which supports mos
t of the ABBYY FineReader Engine API functions through
numerous keys.


MultiProcessingRecognition

Available in
:

C#.


This sample shows the gain in speed when using multiprocessing recognition. Built
-
in multi
processing makes your
application scalable and e
fficient without any efforts from your side.


PDFExportProfiles

Available in
:

C#.


This sample shows the advantages of using PDF export profiles during export to PDF format. These export profiles
contain optimal settings for popular export variants and all
ow you to tune export to PDF with only several
parameters.

ABBYY FINEREADER ENG
INE 10 SYSTEM REQUIR
EMENTS



PC with x86
-
compatible processor (1 GHz or higher).



Operating System: Microsoft Windows 7, Windows Server 2008, Windows Vista, Windows Server 2003,
Windows XP, Windows 2000, and 64
-
bit versions of Windows 7, Windows Server 2008, Windows Vista,
Windows Server 2003, Windows XP.



Memory:



for processing one
-
page documents


minimum 400 MB RAM, recommended 1 GB RAM;



for processing multi
-
page documents


min
imum 1 GB RAM, recommended 1,5 GB RAM.



Hard disk space: 800 MB for library installation and 100 MB for program operation plus additional 15Mb
for every processing page of a multi
-
page document.



100% TWAIN
-
compatible scanner, digital camera, or fax modem


for scanning only.



Video card and monitor (min. resolution 1024*768


for pattern training, dictionary editing, scanning
with a GUI displayed).



Keyboard, mouse or other input device.



The following registry branches should be accessible from the workstation
:



"HKEY_CURRENT_USER
\
Software
\
ABBYY
\
SDK
\
10
\
FineReader Engine"


full control.



"HKEY_CURRENT_USER
\
Software
\
ABBYY
\
SDK
\
10"


full control for installation only.



"HKEY_LOCAL_MACHINE
\
Software
\
ABBYY
\
SDK
\
10"


full control for installation only.



The following fol
ders should be accessible from the workstation:



Folder with ABBYY FineReader Engine binary files


access for reading.



%TEMP% folder


full control access.



%ALLUSERSPROFILE%
\
Application Data
\
ABBYY
\
SDK
\
10
\
Licenses


full control access.



%ALLUSERSPROFILE%
\
Ap
plication Data
\
ABBYY
\
SDK
\
10
\
FineReader Engine


full control access.



The following components should be installed:



Microsoft Internet Explorer 5.0 or higher.



If your application uses pattern training, dictionary editing, scanning with a GUI displayed,
Windows
Common Controls must have version 5.80 or later and Rich Edit Control must have version 3.0 or later.

ABBYY SDK 10 LIC
ENSE SERVER SYSTEM R
EQUIREMENTS



PC with x86
-
compatible processor (1 GHz or higher).



Operating System: Microsoft Windows 7, Windows

Server 2008, Windows Vista, Windows Server 2003,
Windows XP, Windows 2000, and 64
-
bit versions of Windows 7, Windows Server 2008, Windows Vista,
Windows Server 2003, Windows XP.



25 MB of free hard
-
disk space.



The folder %ALLUSERSPROFILE%
\
Application
Data
\
ABBYY
\
SDK
\
10
\
Licenses must have full control access.

INSTALLATION

There are two types of the product installations:



Developer. Used for software development. To set up the product one should run "Setup.exe" from
distribution CD (or its copy).



Runtime.

Used for developed software distribution. To set up runtime copy of the product one should
copy required files and register certain libraries or run "Setup.exe" in command line mode with proper
options.

Please read the Installation Guide document distribu
ted with the CD for details.

SUPPORTED RECOGNITIO
N LANGUAGES


OCR LANGUAGES

1)

With dictionaries:

a.

45 common (default) languages including

i.

6 programming languages

ii.

Chemistry

iii.

Digits

b.

10 additional languages including

i.

Arabic

ii.

Japanese

iii.

Hebrew


iv.

Thai

v.

FR XIX

1.

Old
English

2.

Old French

3.

Old German

4.

Old Italian

5.

Old Spanish

vi.

Vietnames

c.

55 in total

2)

Without dictionaries:

a.

138 common (default) languages including:

i.

4 artificial languages

1.

Esperanto

2.

Ido

3.

Inerlingua

4.

Occidental

ii.

2 special languages (included by default if corresponding

text type is chosen)

1.

CMC7

2.

E13B

b.

6 additional languages

i.

Chinese Simplified (PRC)

ii.

Chinese Traditional (Taiwan)

iii.

Korean

iv.

Hangul (under Korean Add
-
On)

v.

Yiddish (under Hebrew Add
-
On)

vi.

Latvian Gothic (under FineReader XIX Add
-
On)

c.

144 in total

3)

Overall count is 199

a.

183 are common and included in Runtime Professional

b.

16 are included in Add
-
Ons:

i.

Arabic

ii.

5
hieroglyphic languages (CJK)
-

Chinese Traditional (Taiwan), Chinese
Simplified (PRC), Japanese, Korean, Hangul (Korean)

iii.

6 old languages (FineReader XIX)

iv.

Thai

v.

2 Hebrew

languages
-

Hebrew and Yiddish

vi.

Vietnamese


ICR LANGUAGES

1)

29 with dictionary support, new are:

1

Estonian

2

Indonesian (previously was supported w/o dictionary)

3

Slovenian (previously was supported w/o dictionary)

4

Old French

5

Old German

6

Old Italian

7

Old Spanish

2)

84 without dictionary support, 2 have got dictionary support, 2 were miscounted and new are:

1

Buryat

2

Fijian

3

Ganda

4

Hawaiian

5

Kazakh

6

Kirghiz

7

Luba

8

Maori

9

Mongol

10

Nenets

11

Samoan

12

Serbian (Latin)

13

Sotho

14

Tswana

15

Tuvinian

16

Uighur (Latin)

3)

Overall count is 113

1

109 are inclu
ded in Data Capture add
-
on in Runtime Professional license.

2

4 are additionally included in Data Capture add
-
on in Runtime FineReader XIX license.

SUPPORTED BARCODE TY
PES

1D BARCODES

1.

Codabar


2.

Code 128

3.

Code 39

4.

Code 93

5.

EAN 8

6.

EAN13

7.

IATA 2 of 5

8.

Industrial 2 of 5

9.

Interleaved 2 of 5

10.

Matrix 2 of 5

11.

Patch

12.

PostNet

13.

UCC
-
128

14.

UPC
-
A

15.

UPC
-
E

1D BARCODES WITH CHE
CKSUM

1.

Code 39

2.

Interleaved 2 of 5

3.

Codabar

1D BARCODES WITH SUP
PLEMENTAL

1.

EAN 8

2.

EAN 13

3.

UPC
-
E

2D BARCODES

1.

PDF417

2.

Aztec

3.

DataMatrix

4.

QR Code

SUPPORTED
TEXT TYPES, WRITING
STYLES AND FIELD MAR
KING TYPES

OCR TEXT TYPES

1.

Normal

2.

Fax

3.

Typewriter

4.

Matrix

5.

OCR_A

6.

OCR_B

7.

MICR_E13B

8.

MICR_CMC7

9.

Fraktur/Gothic (available only under FineReader XIX add
-
on)

FIELD MARKING TYPES

1.

Simple Text

2.

Underlined Text

3.

Text in Frame

4.

Grey B
oxes

5.

Char Box Series

6.

Simple Comb

7.

Comb in Frame

8.

Partitioned Frame

SUPPORTED IMPORT AND

EXPORT FORMATS

SUPPORTED EXPORT FOR
MATS

1.

RTF

2.

Microsoft Office file formats:

a.

DOC/DOCX

b.

XLS/XLSX

c.

PPTX

3.

PDF file formats

a.

PDF

b.

PDF/A both 1b and 1a

c.

MRC (Mixed Raster Content) for

both PDF and PDF/A

4.

HTML

5.

TXT/CSV

6.

ABBYY XML

KNOWN ISSUES AND WOR
KAROUNDS



CommandLineInterface Sample has Bugs

ABBYY FineReader Engine 10 has many changes in API responsible for setting analysis, recognition,
synthesis, and export parameters. That leads to c
ertain changes in the CommandLineInterface sample.
Unfortunately, not all of them were found and fixed. Nevertheless, any developer can do that by himself.
In addition, we will continue testing of the sample and we are going to free it from bugs by the
mai
ntenance release.



FRECOMWrapper Sample does not Work on 64
-
bit Systems

It is known that FRECOMWrapper sample does not work on 64
-
bit systems. The fix is scheduled for a
future maintenance release.



PCM_ForceToGray has no Effect

PCM_ForceToGray values has no

effect when set up in a PDF export profile, e.g. color images remain
color.



PDFExportParams::ResolutionType = PRT_Exact has no Effect

PDFExportParams::ResolutionType = PRT_Exact

has no effect being set up in a PDF export profile, e.g.
source image resolution is changed to defined by a chosen PDF export profile.



DA Can Consume Too Much RAM

We know a document consisting of 140 pages where automatic analysis can consume more than 1.
5Gb of
RAM. The fix is known but it can be included only in a future maintenance release.



DSPJnsertEmptyParagraphsForBigInterlines = True has no Effect

It is known that DSP_InsertEmptyParagraphsForBigInterlines = true has no effect, e.g.

recognition result

contains no empty paragraphs in place of big interlines.



DSP_DetectDropcaps = False has no Effect

It is known that DSP_DetectDropcaps = false has no effect, e.g. synthesis still detects drop caps.



HTMLSynthesisMode = HSM_PlainText has no Effect



It is kno
wn that HTMLSynthesisMode = HSM_PlainText has no effect, e.g. text formatting remains during
export.

R2


PART NUMBER, BUILD N
UMBER

Part#


681/
17

Build#


10.0.
5
.
166

LIST OF CHANGES

NEW FEATURES AND IMP
ROVEMENTS

SINGLE
-
PAGE DOCUMENTS BATCH

PROCESSING

With this release SDK is presenting a special interface for processing single
-
page documents in parallel
-

IBatchProcessor. A processor object implementing this interface converts input images into the recognized pages
(FRPage objects). When processor is i
nitialized, asynchronous recognition processes are invoked and configured.
Then the processor takes image files from a custom image source. For each page of the image file a new processing
task is created, and this task is passed to one of the recognition
processes. If all the tasks for one file have been
passed for processing, but not all of the recognition processes are occupied, the next image file from the image
queue of the source is taken and passed for processing. This is done until the first image p
age has been converted
and passed to the user. Pages are returned to the user in the order they have been taken from the image
source.“Multi
-
core Batch Recognition” sample is added to the Code Samples Library to demonstrate how the new
API works and how ef
ficiently it utilizes multi
-
core architecture.

ADDITIONAL XML EXPOR
T FORMATS

This release includes several new XML export formats announced earlier and already added to the licensing. All of
them almost have no tuning parameters. Some parameters are planne
d to be added in the next maintenance
release. Customers feedback on what parameters are missed will be very appreciated.

ALTO

ALTO

(
Analyzed Layout and Text Object
) is a XML Schema that details technical
metadata for describing the layout
and content of physical text resources, such as pages of a book or a newspaper. It most commonly serves as an
extension schema used within the
Metadata Encoding and Trans
mission Schema (METS)

administrative metadata
section
. However, ALTO instances can also exist as a standalone document used independently of METS.

This
format was specificall
y developed for
IMPACT

project.


EPUB

EPUB

(
short for electronic publication
) is a free and open e
-
book standard by the International Digital Publishing
Forum
(IDPF). Files have the extension .epub. EPUB is designed for reflowable content, meaning that the text
display can be optimized for the particular display device used by the reader of the EPUB
-
formatted book. The
format is meant to function as a single for
mat that publishers and conversion houses can use in
-
house, as well as
for distribution and sale.

It supersedes the Open eBook standard.

There is no API for tuning in this release, only
ExportFileFormatEnum value FEF_EPUB allowing saving recognition data i
n this format.

FB2

FictionBook (
FB2
) is an open XML
-
based e
-
book format, which originated and gained popularity in Russia. It is
supported by e
-
book readers such as FBReader, AlReader, Haali

Reader, STDU Viewer and Okular under Linux. The
FictionBook files have the .fb2 filename extension. The FictionBook format does not specify the appearance of a
document; instead, it describes its structure. For example, there are special tags for epigraph
s, verses and
quotations. All the ebook metadata, such as the author name, title, and publisher, is also present in the ebook file.
Hence the format is convenient for automatic processing, indexing, and ebook collection management. This also
allows automat
ic conversion into other formats. There is no API for tuning in this release, only
ExportFileFormatEnum value FEF_FB2 allowing saving recognition data in this format.

LOW MEMORY MODE

SynthesisParamsForDocument::LowMemoryMode flag is added to API. It specif
ies whether low memory mode
should be used during document synthesis. If you set this property to TRUE, ABBYY FineReader Engine will try to
use no more than 600 Mb of memory during document synthesis by loading fewer pages to memory
simultaneously. However
, document synthesis may slow down and the quality may slightly deteriorate. By default,
the value of this property is FALSE.

SAVING AND LOADING F
RDOCUMENT FROM A FOL
DER

IFRDocument::SaveToFolder and IFRDocument::LoadFromFolder methods are added to API. Th
is pair of methods
allows to save and load FRDocument object like a document in FineReader desktop product.

OPENING IMAGES FROM
MEMORY

IEngine::OpenImageFileFromMemory and IFRDocument::AddImageFileFromMemory methods are added to API.
They allow opening of
an image file from memory in the following compression formats:



BMP;



DCX;



JPEG;



PCX;



PNG;



TIFF (uncompressed, CCITT, JPEG, Jpeg2K, LZW, ZIP);



Jpeg2K;



GIF;



JBIG2.

SOURCE IMAGE PATH

IFRPage::SourceImagePath properety

was added to API. It specifies the full path to the source image of the page.

WRITING METADATA TO
XMP

IPDFExportFeatures::WriteMetaDataToXmp parameter was added to API. This property is FALSE by default.

It
specifies whether document content information s
hould be written to XMP data of the PDF file. The metadata tags
written to XMP are those from the
Dublin Core set
.

Extensible Metadata Platform (XMP) is a technology, which
allows embedding metadata

directly into a file. It simplifies content information extraction for third
-
party
applications. Normally, document content information is written to the properties of a PDF file. For PDF/A files
document content information is always duplicated to XMP in

compliance with the standard.

FULL VERSION OF IPDF
EXPORTPARAMS

This release presents IPdfExportParams as it was planned for the R1 GM. Briefly it has got the following new
interfaces:



IPdfExportFeatures



IPdfEncryptionInfo



IPdfPaperSizeParams



IDocumentContentInfoWritingParams



IPdfPictureCompressionParams



IPdfMrcParams

XLSX: WRITING EACH D
OCUMENT PAGE TO A SE
PARATE WORKSHEET

XLExportParams::OnePagePerWorksheet property is added to API. It specifies if each page of the document is to be
exported
to a separate worksheet of the XLSX file. This property is FALSE by default. The property may be useful in
the following scenarios:



Combining several one
-
page tables in one XLSX document.



Preserving document data distribution between document pages while e
xporting to XLSX format.

FIXED BUGS

During the last maintenance multiple reclamations were fixed. The most important with a short description is
listed below.



HD 219981


Access violation. The instruction at 0x1F94D95E (c:
\
program files
\
abbyy sdk
\
10
\
finere
ader
engine
\
bin
\
recpage.dll) referenced memory at 0xFFFFFFFC. The memory could not be written
.



HD 204321, HD 221067


No ability to specify code page for 2D barcodes.



HD 213779


IPE “c:
\
builds
\
11.0.2
\
0
\
textlayout
\
text
\
recognizedcharparamsarray.cpp, 1004” if
DocumentInfo is omitted for single
-
page recognition API (IEngine).



HD 217907


IPE ".
\
Src
\
ParagraphExtractor.cpp, 588" while using user trained patterns.



HD 222521


Image p
art
s that are not inside modification rectangles but lay inside the bounding rectangle
are filled in with
black

color when WriteRectImage is called.



HD 221897


FRDocument.PlainText.Text contains no data from table blocks.



HD 219167


Hardware license disappe
ars from License Manager after USB
-
key reconnection.



HD 221685


IPE ”c:
\
builds
\
11.0.2
\
0
\
image
\
libraries
\
rle
\
src
\
regionn.cpp, 1968” during image analysis
using “Text Extraction” profile.



HD 216847


FREngine.dll is missing after runtime installation.



HD 16
6617, HD 205288


Incorrect table analysis.



HD 209678


Empty block is read
-
only.



HD 208010, HD 209476


Bad recognition result. Partially fixed.

KNOWN ISSUES AND WOR
KAROUNDS



Incorrect hardware protection work while a few instances of FREngine try to
access the dongle
simultaneously

Sometimes a protection in the iKey dongle reacts unexpectedly while a few instances of FREngine try to
work with the same dongle simultaneously (
HD 219981
). The problem appears under Windows 2008 R2
x64.

We suppose the pr
oblem is in the dongle itself .



ReportRecognizerTip event isn’t fired for AnalyzeAndRecognizePages

If a document is recognized in parallel mode then ReportRecognizerTip event isn’t triggered during
AnalyzeAndRecognizePages function call.



CJK auto page orie
ntation issues

CJK auto page orientation in the Engine is based on a quality of recognition in all four directions. Since
Asian scripting allows writing in both vertical and horizontal directions the current algorithm has an
obvious drawback and can make m
istakes.

Some improvements are expected in SDK 10 but significant changes are planned only for SDK 11. That is
due to architectural limitations.



Hardware protection under Windows 7 VM



During product release testing cycle we found out that Dongles (iKeys) a
re unavailable if Windows 7 is run
on Virtual Machine PC. On real PC Dongles work fine.



Loss information about skew angle



During Document Analysis a current image is replaced by deskewed one. The information about skew
angle after this process is missed (
HD 214967).


Workaround: to get the skew angle before Document Analysis is started and keep it for the further use.




Error "A drawing error occurred" in Adobe Acrobat v.7



The error occurs on some output PDF files with JBIG compression only.

Workarounds:

avoid using JBIG compression for output PDF;

use Adobe Acrobat version 9 for opening such PDFs.

R2
UPDATE


PART NUMBER, BUILD N
UMBER

Part#


681/
20

Build#


10.0.
5
.
178

LIST OF CHANGES

NEW FEATURES AND IMP
ROVEMENTS

No new features added since the R
2

release.

FIXED BUGS

During the last maintenance multiple reclamations were fixed. The most important with a short description is
listed below.

Access violation while freeing Engine objects via single
-
page API. This error came up after certain changes in
i
nternal memory management.

R3



PART NUMBER, BUILD N
UMBER

Part#


681/24

Build#


10.0.7.157

LIST OF CHANGES

NEW FEATURES AND IMP
ROVEMENTS

VISUAL COMPONENTS

This release includes one of the most expected and desired ABBYY FineReader Engine features


Visual
Components module which provides graphical user interface similar to the interface of ABBYY FineReader. With
the Visual Components you can easily create an

application for viewing and editing images, editing and verifying
recognized text and monitoring document processing.

ABBYY FINEREADER ENG
INE AND CLOUD TECHNO
LOGY

We are glad to announce that since new version FineReader Engine 10 has technical backgroun
d for integrating
with Windows Azure.

New section added to the Developer’s Help: Deploying ABBYY FineReader Engine on Windows Azure. This section
provides step
-
by
-
step instruction on how to deploy an application, which uses ABBYY FineReader Engine library
, on
Windows Azure.

The availability of this feature is controlled by the Licensing Subsystem. It is registered as a property in section
“Environment limitation” of License Properties.

Please, note new GetEngineObjectEx function added to API. This functi
on is used to get a pointer to the IEngine
interface. The only difference from the GetEngineObject function is that GetEngineObjectEx function takes as input
parameters the path to the folder with user
-
specific data and the path to the folder with temporar
y files. This
function is useful in the situations when default folder with user data is inaccessible (e.g. it is read
-
only on
windows Azure).

RUNTIME LICENSE EMUL
ATION

It is possible now to use a Developer's license in the Runtime emulated mode. That allo
ws testing your application
which uses ABBYY FineReader Engine 10 library.

NEW BINARIZATION

PrepareImageMode::UseFastBinarization flag is added to API.

With this property ON ABBYY FineReader Engine will use algorithms for fast image binarization

during preparation.
This feature speeds up binarization however its quality may deteriorate.

JAVA SAMPLES

Hello sample for Java has been added.

This Hello sample demonstrates how to use ABBYY FineReader Engine in Java within JNI (Java Native Interface).

It
includes a C++ “HelloJNI” project with a sample JNI
-
wrapper implementing the most essential methods of
FRDocument object.

NEW MARKETING SAMPLE
S

There are several samples showing how to use Engine in different scenarios added to the new release:



“Visu
alComponents” is a helpful sample source code for working with Visual Components module.



“Engine Predefined Processing Profiles” shows how to use the predefined profiles to configure FineReader
Engine.



“Camera OCR” is devoted to the photographed documents
and the appropriate FineReader Engine tools
for improving the recognition quality.



“Image Preprocessing” visually shows different preprocessing tools in work.

NEW INTERFACE LANGUA
GE

New release includes Greek interface language supported.

SAVING ORIGINAL

PAGE SIZE IN PDF EXP
ORT MODE

IPDFPaperSizeParams:: SetPageSizeByOriginalImageSize property is added to API.

It specifies if the page size must be equal to the original image size during export of the recognized text into PDF
format.

SEPARATE PAGES IN HT
M
L EXPORT MODE

HTMLExportParams:: SeparatePages property is added to API.

If this property is TRUE and several pages are exported to HTML format, <HR> tag is inserted between pages,
which makes the browser draw a horizontal rule. In this case, the logical s
tructure of the document is not retained.

THE SIGNATURE OF FIN
DPAGESPLITPOSITION M
ETHOD HAS CHANGED

The number of input parameters of FindPageSplitPosition (IDocumentAnalyzer::FindPageSplitPosition;
IFRPage::FindPageSplitPosition) has been changed. The Tex
tOrientation object is passed as an input parameter.

FIXED BUGS

During the last maintenance cycle much reclamation were fixed. The most important with short
descriptions are listed below.




HD 244309


E_OUTOFMEMORY on IFRPage::Analyze
.



HD 242913


Unhandl
ed exception at 0x2eac4fc4 (DocumentAnalysis.Segmentation.dll) in
Reclamation.exe: 0xC0000094: Integer division by zero ( .
\
src
\
ComParametersConverter.cpp, 86)
.



HD 238795


Landscape orientation detected instead of original portrait.



HD 242367


Soft
hyphens are wrongly replaced by hyphen
-
minus in the resulting PDF text layer.



HD 231507


Developer’s Help section “
Working with the LicensingSettings.xml File
” is updated
.



HD 226429


Developer’s Help section “TextOrientation Object” is updated.



HD 230962



Distribution list is not the same as bin folder contents.



HD 221685


IPE ”c:
\
builds
\
11.0.2
\
0
\
image
\
libraries
\
rle
\
src
\
regionn.cpp, 1968” during image analysis
using “Text Extraction” profile.



HD 240810


Interleaved 2 of 5 barcodes
processed
incorrectly.



HD 241572


Symbol
"

" not recognized if Japan recognition language chosen
.



HD 224227


“ABBYY Licensing Service is unavailable: The RPC server is unavailable” if refreshing License
Manager while multipage document is being processed on Win 20
00
.



HD 226945

-

Inconsistent recognition parameters: properties Perform Training and TextType are set to
incompatible values.



HD 233409


Splitted symbols with
txtBlock.RecognizerParams.OneLinePerBlock = true and
txtBlock.RecognizerParams.OneWordPerLine =
true
.



HD 234388



" Execution of parallel task failed: Access violation. The instruction at 0x1D6DD370
(d:
\
abbyy
\
fre10
\
bin
\
recpagehp.dll) referenced memory at 0x0000001C. The memory could not be read."



HD 236572


Incorrect table analysis.



HD 209678



Dev
eloper’s Help section “ABBYY FineReader Engine Distribution Kit” and
FREngine10_Distribution.csv are modified: Cjk.* and Arabic.* files in Distribution are not mandatory.



HD 226429

-

ImageDocument
-
>get_ImageRotation returns RT_Counterclockwise when image i
s Clockwise
rotated.



HD 226373


When
FRDocument API is used ImageRotation is always “RT_NoRotation”
.



HD 228364
,
HD 228047
,
HD 224128
,
HD 226384
,
HD 227306
,
HD 227082
,
HD 226597



Developer’s Help
sections Guided Tour and FAQ are updated with the informati
on about
DocumentInfo object
.



HD 221312


Bad document structure when creating new Docx adding existing docx files created by
FREngine.



HD 238336



Developer’s Help sections referring to
%ALLUSERSPROFILE% folder for the installation

are
updated with
%Progr
amData% folder for appropriate OS.




HD 238795


DetectTextColor affects on the export quality when exporting to PDF format
.



HD 227357


Incorrect
CharConfidence

value exported to XML file.



HD 237962


ParagraphParams values are not filled for mixed
-
langu
age texts when English+Hebrew
selected for recognition
.



HD 240119



E_OUTOFMEMORY on FRDocument::AddImageFile.



HD 237980


OutputDebugString dumping into debugger log.



HD 218664


Barcode values are not written to the exported XML file.



HD 235713


FontMo
de property added to PDFExportParams
.



HD 232321



Colors on image are inverted when exporting to PDF ImageOnly after opening and synthesis
of the original image.



HD 225832


Developer’s Help sections “SynthesisParamsForPage Object” and
“SynthesisParamsForD
ocument Object” are updated with
DetectFontFormattingAtPageLevel description
.



HD 230007


Data
-
Matrix barcode is not recognized
.



HD 219981


License storage (Dongle 57D4
-
64030005 on GHSR09000016) was removed from ABBYY
Licensing Service.



HD 225959

-

R
untime installation requires serial number.



HD 230096


IPE: .
\
Src
\
AnalyzeDocumentManager.cpp, 157 on BatchProcessingRecognition sample.



HD 222452
,
HD 216668



Error during document processing. Unsupported font Arial when exporting to
PDF
.



HD 228739


IPE:

.
\
src
\
PlainText.cpp, 123
.



HD 202214


IPE: .
\
Src
\
TableCellsStructure.cpp, 344.



HD 223505


IPE: .
\
Src
\
PDFAdjuster.cpp, 3777 when exporting to Searchable PDF
.



HD 223919


Incorrect
RotationType
.



HD 223505


IPE: .
\
src
\
ProcessorsPool.cpp, 104.

KNOWN ISSUES AND WOR
KAROUNDS



Error “One or more arguments are invalid” in XLS/XLSX
export

HD 240747

The issue occurs if PaperSize parameter of XSLExportParams is set to XLPS_A2. The A2 size of XLS/XLSX
format is not supported by current FREngine 10 versio
n but it should be fixed in next versions.



Access Violation error when running FREngine

During our tests an AV error is noticed to arise occasionally in FineNet library. E.g. it happens in case of
loading FRDocument that has been saved before: "Unhandled e
xception at 0x2e3e5924 (FineNet.dll) in
Reclamation.exe: 0xC0000005: Access violation reading location 0x0000007e."


Visual components issues

As this is the first version of Visual Components module in FREngine 10, it still has some issues to be fixed
and
improved. The most common problems concern buttons and menu items control, text viewer etc but
below there are some most important issues to note:



IPE (
c:
\
build
\
ll.0.5
\
0
\
interfaces
\
implementations
\
inc
\
notifyimpl.h, 62)

can occur when processing several
do
cuments simultaneously in VC and open one more document.



The export modes are not working in Saving dialog: all the pages are saved in one file.



“Merge Rows” in ImageViewer menu is not actual: it does not merge table rows.

As these issues are not critical

we release this version and continue improving Visual Components for the
next ones.



Vertical text is not captured if page orientation is not detected

In case of the multi
-
page document with differently oriented pages text sometimes it is necessary to save

the orientation in the resulting file. At present to recognize rotated images the orientation has to be
detected (DetectOrientation = TRUE). As a result these pages are exported rotated to portrait orientation.
In case you do not detect orientation (Det
ectOrientation =FALSE) the text on the rotated images is not
recognized properly. Though this it is not a bug it shoud be improved in future.



Document structure is not kept in HTML export



HD 230061

In case of saving the recognition results in HTML format
the structure of the original document is lost (e.g.
text columns). Though it has not been planned before, the issue has been fixed after this case and now
requires additional testing.



Barcode recognition downgrade



Some barcode types (
EAN13, EAN8
) are re
cognized with worse accuracy than in FREngine 9.0 version.




Text color (RTF export) and “DocumentArchiving” profile



When using DocumentArchiving profile text color is lost in exported RTF files.

R3

UPDATE


PART NUMBER, BUILD N
UMBER

Part#


681/28

Build#


10.0.7.162

LIST OF CHANGES

NEW FEATURES AND IMP
ROVEMENTS

No new features added since the R3 maintenance release.

FIXED BUGS

During the last maintenance cycle much reclamation were fixed. The most important with short descriptions are
listed below.



HD 249923


Incorrect FineReader Engine behavior when switching
SharedCPUCoresMode with single
-
cored license.




HD 247120



IPE:.
\
Src
\
WordExporter
\
WordExporter.Utils.cpp, 322” when exporting to rtf with
IRTFExportParams::BackgroundColorMode = BCM_DontSave
.



HD 248274


“IPE: c:
\
build
\
11.0.trunk
\
0
\
ocr
\
recpage
\
fragmrec
\
normhyp.cpp, 373” when processing
document using CompoundTextLanguage with Japanese language only.



HD 250025


“IPE: .
\
Src
\
PdfContents.cpp, 416” when exporting to PDF.




HD 249747


“IPE:

d:
\
b
uild
\
11.0.5
\
0
\
ocr
\
shared
\
rectools
\
globaltranslationtable.cpp, 191

.

KNOWN ISSUES AND WOR
KAROUNDS

HP 249923

This realize includes fixes that allow to implement workaround for those who is using SharedCPUCoresMode with
one
-
core license. In that case the En
gine object must be initialized the following way:

// Call the GetEngineObject function with null as developerSN int hresult = GetEngineObject(null, null, null, ref
engine); Marshal.ThrowExceptionForH R(hresult);

// switch to SharedCPUCoresMode

engine.Mult
iProcessingParams.SharedCPUCoresMode = true; // initialize licensing

FREngine.lLicense license = engine.Licenses.FindLicense(developerSN); engine.SetCurre
ntLicense(license,
developerSN);

R4


PART NUMBER, BUILD N
UMBER

Part#


681/31

Build#


10.0.8.109

LIST OF CHANGES

NEW FEATURES AND IMP
ROVEMENTS

VC: TITLE PROPERTY

Following the request
HD 252106

new ‘Title’ property was added to ImageViewer, DocumentViewer and
TextEditor visual components.
This allows changing and hiding titles of those components.

‘VERSION9COMPATIBILI
TY’ PRE
-
DEFINED PROFILE DESC
RIBED IN RESOURCES

This release contains resources for ‘Version9Compatibility’ pre
-
defined Engine profile. Previous releases have this
profile ‘empt
y’, i.e. equal to ‘Default’ or ‘DocumentConversion_Accuracy’.

‘USING ABBYY FINEREA
DER ENGINE IN C (OBJ
ECTIVE
-
C)’ ARTICLE IN THE H
ELP FILE

HD
242862

Those who develop in C (Objective C) languages have got a guideline in the Engine documentation (please see
‘Using ABBYY FineReader Engine in C (Objective C)’ article) facilitating Engine usage.

NEW EPUB EXPORT PARA
METERS

HD
246952

IEbookExportParams interface is extended with the following new properties:



FontAttributesMode
. Specifies which font attributes are
to be written.



MetaDataWritingParams
. Specifies if the author, subject, title, and keywords of the document should be
written.



PictureFormat
. Specifies the image format to be used.



PictureJpegQuality
. Stores the value in percentage points of the JPEG quali
ty for color pictures.



PictureResolution
. Stores the value of picture resolution in dpi, which is used for exporting pictures.



TreatFirstPageAsCover
. Specifies whether the first page should be treated as a cover.

‘ITEXTBLOCKANALYSISP
ARAMS::SKEWCORRECTIO
NMO
DE’ HAS NEW DEFAULT
VALUE

ITextBlockAnalysisParams::SkewCorrectionMode has got new default value
-

SCM_AccordingToPage. This value
instructs Engine to do skew correction on a block level same way as it is done for a page.

IIMAGEDOCUMENTS::GRA
YIMAGE()

Start
ing from this release IImageDocument interface gets new property ‘GrayImage’. It returns a gray layer of an
image (image in a grayscale colority) as IImage object.

BUILT
-
IN COM WRAPPER AROUN
D IENGINE FOR SERVER

AND 64
-
BIT HOST APPLICATION
S

This release pre
sents new ways of loading Engine object: as in
-
proc or out
-
proc COM object. This gives an ability
easily organize multithreaded server application with Engine instances pool; or develop 64
-
bit native application
that communicates with 32
-
bit Engine object
via COM transparently.

Please see details and guidance in the following Help articles:



‘Different Ways to Load the Engine Object’;



‘Using ABBYY FineReader Engine in Multi
-
Threading Server Applications’.

PRE
-
CREATED JAVA WRAPPER

To facilitate development in

Java programming language this release includes Java wrapper around Engine API
(please refer to ‘…
\
Inc
\
Java
\
com.abbyy.FREngine.jar’). That should unlock full Engine API limited earlier by only
‘Hello’ sample in C++ and JNI.

DEBUG MODE IN RUNTIM
E IS ALLOWE
D BY A SPECIAL OPTIO
N IN A LICENSE

Starting from this release we officially announce an option in Licensing that allows debug mode in runtime.

To leverage the feature two things have to be done:



checkmark is checked in a runtime license (find it on ‘Gener
al (FRE, FCE)’ tab in the
License Template Wizard);



‘Protection.Developer.dll’ is distributed instead of ‘Protection.dll’.

TOOLBAR FOR A DOCUME
NTVIEWER

This release presents a Toolbar property of DocumentViewer visual component. It allows forming of a tool
bar with
buttons (items) of the following types:



MI_Options;



MI_LanguageEditor;



MI_ViewDictionaries;



MI_RenumberPages;



MI_Separator;



MI_CustomCommand1…10.

‘USER PATTERN TRAINI
NG UTILITY’ SAMPLE C
ODE IN C#

Samples collection was appended by a new sample in
C#: User Pattern Training Utility. It allows you training of
patterns to unusual fonts using Engine and Visual Components in a way similar to that available in ABBYY
FineReader desktop product. The utility saves patterns in a format compatible with the Eng
ine.

A need in the utility appeared due to changes in ABBYY FineReader 10 project format which was incompatible with
SDK and incorporated patterns along with other document (batch) data. Thus the utility is an alternative for people
who prefer (or like) to

train patterns in FR GUI and then pass them to SDK.


FIXED BUGS

During the last maintenance cycle much reclamation were fixed. The most important with short descriptions are
listed below.



HD 259679


Pages[i].PlainText.SaveToTextFile

saves recognition result of pages with indices smaller than
‘i’
.



HD 228047,
227306



IPE during de
-
initialization.



HD 261530


IPE ‘D:
\
build
\
11.0.5
\
awl
\
inc
\
windowaw.inl, 295’ appears in a log for subsequent calls for
scanning multipage documents.



HD 24851
1


PDF file is converted to blank ImageDocument.



HD 257004,
248853



Two
-
digits page numbers are written in two lines.



HD 248592,
262040



The installer in runtime mode accepts SNs for ABBYY FineReader 10 Professional
Edition.



HD 219981


Access Violation

in IEngine::RecognizePage() call.



HD 246585


Text and table losses in PDF to RTF conversion scenario.



HD 255254


IPE ‘.
\
DA7
\
ImageObject.cpp, 794’ on certain image during document analysis.



HD 253031


Crash on IFRDocument::Process() called after canceli
ng previous process in a callback.



HD 253031


Incorrect external names of built
-
in
English_Address and Czech_Address languages for
German message language.



HD 247121


RTFExportParams::PageOrientation property has no effect.



HD 248882


PictureCompression
Params::ColorFormat property has no effect.



HD 248892


Incorrect description for ‘Extended Character Info’ licensing module in the Help file.



HD 248854


Bold
‘Calibri’ font is chosen instead of normal.



HD 253122


The Help states that ‘
CCITT3FAX’
compression is available for writing.



HD 249747


IPE ‘
d:
\
build
\
11.0.5
\
0
\
ocr
\
shared
\
rectools
\
globaltranslationtable.cpp, 191’ appears in a log
file.



HD 249923


‘WARNING: no suitable vouchers found’ appears in a log file for a standalone license and
correc
t Engine processing parameters set.



HD 222640


IPE ‘
c:
\
builds
\
ll.0.3
\
0
\
ocr
\
recpage
\
fragmrec
\
normhyp.cpp, 366’ during document analysis.



HD 246660


Wrong page orientation detection.



HD 246634


Incorrect name of ‘
FlexiFormsDA
’ property appears in a log fi
le.

KNOWN ISSUES AND WOR
KAROUNDS



Pre
-
created Java wrapper has memory leaks

On large image batches (~100 pages) memory leaks are noticeable while using pre
-
created Java wrapper.



IXMLExportParams::WriteNondeskewedCoordinates doesn’t work



Property IXMLExport
Params::WriteNondeskewedCoordinates has no effect.



Error “One or more arguments are invalid” in XLS/XLSX export



HD
240747

The issue occurs if PaperSize parameter of XSLExportParams is set to XLPS_A2. The A2 size of XLS/XLSX
format is not supported by curre
nt FREngine 10 version but it should be fixed in next versions.

R5


PART NUMBER, BUILD N
UMBER

Part#


681/34

Build#


10.0.9.89

LIST OF CHANGES

NEW FEATURES AND IMP
ROVEMENTS

ODT (OPEN OFFICE DOC
UMENT) FORMAT SUPPOR
T

This release presents support for Open
Office document (ODT) export format.

The parameters of a file in this format are tuned through the RTFExportParams object.

JAVA WRAPPER FOR 64
-
BIT APPLICATIONS

This release includes ‘com.abbyy.FREngine.jar’ with two versions of Java wrapper (FREngine.JNI.d
ll): 32
-
bit for x86
architecture and 64
-
bit for x64 architecture. Appropriate DLL is used depending on Java machine bitness.

Hint
. To use FRE from 64
-
bit Java application ‘OutprocLoader’ must be called to get IEngine interface.

IEBOOKEXPORTPARAMS::
WRITEPIC
TURES

IEbookExportParams interface was appended with ‘WritePictures’ property.

It specifies whether pictures should be written in output file. By default this property is TRUE.

VISUAL COMPONENTS EX
TENSIONS

DocumentViewer

component got new event ‘
OnRenumberPage
’. It is called by ABBYY FineReader Engine before
renumbering one page. It delivers to the client the old and the new index of the page. With the help of this method
it is possible to cancel the operation.

DocumentViewer
,
ImageViewer
,
ZoomVie
wer
, and
TextEditor

components got new property
‘ContextMenuEnabled’. It specifies whether the context menu is enabled. If this property is FALSE, context menu is
never displayed. By default, the value is TRUE.

All

components got new event ‘OnOptionsChange
d’. It is called by ABBYY FineReader Engine after the option have
been changed in a component. Options are specified in the Options dialog box, which is displayed if the
MI_Options command is called, or if the Options button is clicked in a dialog box of c
omponent. It informs the
client that the options have been changed.

IREGION::ISEMPTY

IRegion interface got new property ‘IsEmpty’. It specifies whether the region is empty.

Since a region is far more complicated than simple rectangle it requires items iter
ation to check if a region is
empty. At the same time many of FRE methods do not accept empty regions.

FIXED BUGS

During the last maintenance cycle much reclamation were fixed. The most important with short

descriptions are
listed below.



HD 284299, HD 2799
64, HD 287361


An error ‘There is a problem with this Windows Installer package. A
program run as part of the setup did not finish as expected. Contact your support personnel or package
vendor. Action Launch_deinstall_VisualCo
mponentsX.dll’ during SDK de
-
installation.



HD 257004


Unstable synthesis and RTF export result in terms of document layout reconstruction.



HD 240882, HD 249888, HD 285581


Error message ‘Requested functionality is not available’ in case of
PES_Balanced and PES_MinSize PDF export sc
enarios usage without MRC licensing module available. This
is correct behavior. Help file is updated correspondingly.



HD 242862


Help file has no guidance how to compile a host application using GCC.



HD 287754


Incorrect positioning of a table and aligni
ng of data in cells.



HD 278132


Wrong page orientation in output RTF file.



HD 201684


Missing Hiragana and Katakana in output PDF file under OS with Japanese GUI.



HD 199525


Virtual memory leakage.



HD 231157


Error: ‘Runtime Error! Program: C:
\
Program
Files
\
Panasonic
\
OCREngine
\
PnOcrPrcs.exe
R6025
-

pure virtual function call’,
-

during OS shutdown if a host application already called
‘GetEngineObject’ and ‘LoadModules’.



HD 273052


Scanning on Panasonic kv
-
s5055c ends with ‘Fatal scanner error’.



HD 2874
77


Export to PPTX. Last text strings overlap each other.



HD 246585, HD 278132, HD 287233


Export to Word format in PSM_RTFExactCopy mode. Too large page
borders.



HD 275635


Scroll bar in validation form causes application crash.



HD 268262, HD 251164


Character rectangle partially situates outside text block area.



HD 281009


Error: ‘
4440 :Cannot open the file. The file may be corrupted.
’,
-

on a batch of more than 30
images when FRPage::put_Layout method is used.



HD 281347


Incorrect progress reportin
g during call to IExporter::ExportPages.



HD 281347


Incorrect progress reporting during call to IDocumentAnalyzer::AnalyzeAndRecognizePage
with option IPageProcessingParams::De
tectOrientation = true.



HD 280866


Source image is still blocked after releas
ing FRDocument object.



HD 277453


Application hangs after canceling scanning.



HD 284542


IPE: ‘c:
\
build
\
11.0.trunk
\
0
\
fin
eobjects
\
inc
\
map.h, 354’,
-

on opening corrupted JP2 files.



HD 281125


I
PE: ‘.
\
Src
\
PDFAdjuster.cpp, 4359’,
-

on exporting to PDF a
document with manually added
text block.



HD 274046


Skew correction information is lost after page orientation correction.



HD 255254


I
PE: ‘.
\
DA7
\
ImageObject.cpp, 794’,
-

during page analysis.



HD 276972


Error
: ‘Access Violation’,
-

during synthesis wit
h Layout prepared earlier for the same image.



HD 277945


Loosing color of drop caps.



HD 276897


No recognition variants in output XML file.



HD 274543


Incorrect recognition of ‘

’.



HD 242934


Export to RTF, ExactCopy mode. Text situates under an image.



HD 271356


No support for certain type of JP2K image file header.



HD 271111


Can’t detect DataMatrix code on an image.



HD 240747


XLPS_A2 paper size is not supported.



HD 275635


I
PE: ’d:
\
build
\
11.0.8
\
0
\
fineobjects
\
src
\
resourceinfo.cpp, 476’,
-

during
Engine object unload
if ScanManager object was created.



HD 272500


VC, DoCommand. Error while deleting block.



HD 254346


Additional white spaces in output PDF file.



HD 274426


I
PE: ‘.
\
src
\
ExternalObjectsCache.cpp, 54’,
-

when ImageDocument object is pas
sed to the
second FRDocument while the first FRDocument with the same ImageDocument object inside was closed.



HD 271137


Error

‘Unknown COM error. Error code: 0x80040202’ if IEngine object created via
OutprocLoader and Synchronizer.Document function is ca
lled.



HD 272217


I
PE: ‘.
\
Src
\
TextColumnIncutsFinder.cpp, 2805.’.



HD 252106


Lost backward compatibility in VC. No ability of hiding scroll bars in ImageViewer and
DocumentViewer.



HD 252106


No ability of hiding language pane in DocumentViewer.



HD 252106



No ability of hiding property pane in ImageViewer.



HD 269729


I
PE: ‘Src
\
BlockTemplate.inl, 69’,
-

on ImageDocument.CorrectSkew() call if image resolution
was manually changed.



HD 266875


IXMLExportParams::WriteNondeskewedCoordinates parameter doesn’t
work.

R5 UPDATE


PART NUMBER, BUILD N
UMBER

Part#


681/37

Build#


10.0.9.92

LIST OF CHANGES

NEW FEATURES AND IMP
ROVEMENTS

No new features added since the
latest

release.

FIXED BUGS

During the last maintenance cycle much reclamation were fixed. The most
important with short descriptions are
listed below.



HD 289495
,
HD 290431



FREngine.dll
’ is not registered by product installer.



HD 270065


CharConfidence in XML output differs from that got via API.



HD 283009


Duplicating table title in TXT output.



HD
278201


Incorrect table recognition. Fixed partially.

R6


PART NUMBER, BUILD N
UMBER

Part#


681/41

Build#


10.0.10.72

LIST OF CHANGES

NEW FEATURES AND IMP
ROVEMENTS

JAPANESE INTERFACE L
ANGUAGE

This release includes support of Japanese message language.
Now text in all GUI elements, including Visual
Components and Training User Patterns dialogs, can be displayed in Japanese.

The only limitation: special predefined languages for data capture (e.g. Bulgarian address, Croatian personal name)
were not transla
ted to Japanese and are displayed in English.

RECOGNITION OF CHECK
MARKS OF CUSTOM TYPE

This release includes new interfaces and methods, which allow one to train FineReader Engine to recognize
checkmarks of non
-
standard form.

This procedure is described in

detail in the Recognizing Checkmarks section of the Help file.

SUSPICIOUSDISTANCE,
BLACKTHRESHOLD, AMOU
NTOFBLACK PROPERTIES

ARE NOW
AVAILABLE FOR A CHEC
KMARK BLOCK

Previousely hidden SuspiciousDistance, BlackThreshold, AmountOfBlack properties are impleme
nted in this release
and are described in the Help file.

BINARY DATA OF DATA
MATRIX AND QR CODE B
ARCODES

Since this version one can work with binary data of Data Matrix and QR Code barcodes. The procedure of working
with binary data is the same as for othe
r barcodes with binary data (it is described in Working with Barcodes in the
Help file).

RECOGNITION SESSION
CLEANUP

The logic of cleaning of recognition session has been changed. Now recognition session is automatically cleaned
after recognition of each p
age. Therefore recognition in sequantial mode produces the same result as in parallel.
Automatic cleanning can be turned off by using the AutoCleanRecognizerSession property of the
IDocumentAnalyzer interface.

FIXED BUGS

During the last maintenance cycle m
uch reclamation were fixed. The most important with short descriptions are
listed below.



HD 305850



IPE: ‘c:
\
Build
\
11.0.trunk
\
0
\
DocumentAnalysis
\
Inc
\
DocumentAnalysis.Quality.inl, 24.’
-

during Demo.tif processing
.



HD 300563



IPE: ‘.
\
Src
\
RecLinesExtractor
.cpp, 703.’ when processing with "English,Arabic" recognition
language and FlexiFormsDA mode.



HD 298731



An “out of memory” error occurred after about twenty
-
four hours of processing of big jpg
image files (A0 format, 300 dpi, color jpg images, 20 Mb file
s) in VB6.



HD 296803



An error in Java interop, which led to an error during Engine initialization from Java.



HD 295358



IPE: ‘src
\
ExternalEnumConverter.h, 55.’


when calling
IWordRecognitionVariant::get_ModelType. The new
WMT_MRZString constant was add
ed to
WordModelTypeEnum.



HD 293860



Export to RTF format of the document, which contains pages of different sizes (A4,A3), is
supported.



HD 297615



Incorrect LicensingSettings.xml was created during runtime installation with the parameters:
LICENSESRV=Ye
s, IKEYDR=No.



HD 302551



Invalid XML document was created when exporting barcode value with binary data.



HD 303644



QR Code with slightly distored right and bottom borders was not recognized



HD 303593



Incorrect values for LI_Zulu and LI_Yiddish in
Lang
uageIdEnum



HD 302890



At customer’s suggestion,
ICharParams::IsFromSourceContent was added to API. The
property specifies whether the character was recognized or was extracted from the text layer of a source
file. However the property works not perfect.



HD 292363



The value of the fromPos parameter was ignored in the GetRange and Remove methods of
the IText interface.



HD 287233
,
HD 305489



Incorrect page orientation during export to RTF.



HD 281125



If a text block was created manually and an empty para
graph was added to it and then the
document was exported to PSM_RTFExactCopy, the error occurred: IPE
.
\
Src
\
WordExporter
\
WordDocument.Frames.Impl.h, 88.



HD 282909



Incorrect conversion of coordinated between IT_Deskewed and IT_Base.



HD 293557



The Help f
ile was updated in the section of checkmark recognition.

R6 UPDATE


PART NUMBER, BUILD N
UMBER

Part#


681/45

Build#


10.0.10.76

LIST OF CHANGES

FIXED BUGS

During the last maintenance cycle much reclamation were fixed. The most important with short
descriptions are
listed below.



HD 313872



IPE: c:
\
build
\
11.0.trunk
\
0
\
fineobjects
\
inc
\
map.h, 346 during recognition with
IFRDocument::Recognize and default settings + Dutch language.



HD 309198



Incorrect recognition of a table row.



HD 321725



Increased t
ime of export to Word formats due to resaving of image data.

R7


PART NUMBER, BUILD N
UMBER

Part#


681/46

Build#


10.0.1
1.98

LIST OF CHANGES

NEW FEATURES AND IMP
ROVEMENTS

EDITABLE CONTEXT MEN
U OF VISUAL COMPONEN
TS

The Context menu of Visual Component in FREngine 9.0 was available for editing. Initially in FREngine 10 this
functionality was removed. Because of customer requests it was restored in this version. Implementation is
different from that of FREngine 9.0. Se
e the description of the new
IPopupMenu

interface and the
Customizing
the Context Menu and Toolbar of a Visual Component

section in the Help file.

UPDATED EXPORT TO AL
TO FORMAT

ALTO export format received a number of improvements in this version:



Location
of the schema is written in the file.



Incompatibilities with the schema have been corrected.



The page is divided into print space and margins.



Base lines are written to the files



Character coordinates can be written relative to the source image. This is de
fined by the
IAltoExportParams::WriteNondeskewedCoordinates

property.

RUNTIME DEINSTALLER

HD #321320
,
HD #317932


HD #269260



Runtime deinstaller has been added thanks to customer requests.

The new key
deinstallruntime=yes/no
has been added. If you run
se
tup.exe /qb /v installDir=<path to the
FREngine installaltion folder> deinstallruntime=yes
, the license service and all FREngine folders and registry keys
will be deleted.

The key has not yet been described in product documentation. The description will be

added in the next release.

HOT KEYS CAN BE SPEC
IFIED FOR CUSTOM COM
MANDS IN VISUAL COMP
ONENTS

When registering a new custom command using the
RegisterUserCommand

method of the Commands object, one
can specify a hot key for
the command.

FIXED BUGS

During t
he last maintenance cycle much reclamation were fixed. The most important with short descriptions are
listed below.

Image Opening and Preprocessing



HD #321725
,
HD #313037


Temporary Image Document folders were not removed from the
%temp%
\
ABBYY FineReader
Engine 10 folder.



HD #308285



If a large PDF file was opened and it had a damaged page, the whole file could not be
opened.



HD #322802



When a TIFF image saved in FREngine 8 containing an image and its preview was opened in
FREngine 10 the GetNumberOfPag
esInImageFile method returned 2. If the same image was added to
FRDocument, the number of pages in the document was 1. Now the GetNumberPfPagesInImageFile
method returns the same number of pages as the number of pages in FRDocument (1).

Analyses, Recogniti
on, and Synthesis



HD #320192



Access violation error when a document was analyzed with recognition languages:
"ChinesePRC,English".



HD #267143



Incorrect alignment detection for some blocks.



HD #301672



The names of the fonts in the FontCache folder con
flicted with the names of system fonts.
To resolve the issue, the names have been changed.



HD #312172



If IPageAnalysisParams::ProhibitModelAnalysis = true and
IObjectsExtractionParams::FlexiFormsDA = true are used simultaneousely, an empty result was pro
duced.



HD #316520



An error during IFRDocument::Synthesize() method call:
IPE: .
\
Src
\
PageTextCluster.cpp,
1168.



HD #316078



An error during analysis of a page using IFRDocument::Analyze IPE: .
\
Src
\
TableTools.cpp,
1128.



HD #312510



An error when
recognizing a document in Swedish with default fonts: IPE:
.
\
Src
\
FontVariant.cpp, 140.



HD #307128

-

E_OUTOFMEMORY error when processing a large document.



HD #302383



Pending of the IFRDocument::Analyze method on the 63d page of a document.



HD #276106



In
correct orientation detection on a page of a multi
-
page PDF.



HD #310501



A table was not found even if a table block was added manually.



HD #307044



An error on a certain image when calling IDocument::Process():
IPE:.
\
Src
\
ComplexPictureSectionCreator.cpp
, 736.

Export



HD #326097



For some document export to RTF ExactCopy with KeepLines parameter set to TRUE results
in an incorrect order of words in paragraphs.



HD #305489



For some pages export in PSM_RTFEditableCopy mode resulted in 2 pages with landscap
e
orientation, when it should be 1 page with portrait orientation.



HD #270065



CharConfidence value extracted from an output XML file and
ICharacterRecognitionVariant::CharConfidence value for the same character could be different. There was
an error in
XML export, it has been fixed.

Visual Components



HD #328111



In Visual Components, if after recognition of a page of some document one call the
command “Analyze table structure”, an error was returned: Internal Program Error:

c:
\
Build
\
11.0.trunk
\
0
\
Documen
tProcessing
\
DocumentProcessing
\
Src
\
BlocksTemplate.inl, 124.



HD #328005



In Visual Components, if a page, which contains a table, was recognized twice, an error was
returned: Internal Program Error:

c:
\
build
\
11.0.trunk
\
0
\
textlayout
\
text
\
recognizedtext.cpp,

71.



HD #314151



the SettingsPath property of Visual Components now is defined by the Engine data folder
specified in GetEngineObjectEx function during initialization.



HD #308785
,
HD #252001



Errors when running a Visual Components project in
Delphi2010.



HD #228064



Only predefined dictionaries were available during verification in TextValidator, but not
custom dictionaries.



HD #300210



Suggestion to add the Has method for Toolbar object.

Barcodes



HD #319860



Some barcodes were not recogniz
ed in a document even if barcode blocks were specified
manually.



HD #312622



Barcodes with white noise on black bars were not recognized.

Documentation



HD #316762



Incorrect description of FREngine.dll registration in the Help file.



HD #321080



Incorrec
t description in documentation: Coordinates of table cells returned via the Left,
top, Right, Bottom properties are not in base grid, but in pixels.



HD #310514



Insufficient information in the Help file: all user patterns are ignored if at least one CJK
language is selected for recognition.

Other



HD #311725



If OutprocLoader was used in
Java
, the created FREngine.exe process was not finished after
the engineLoader.ExplicitlyUnload

method call, even if there are no references on the objects. Calling
garbage collection was not helpful too.



HD #277002



When running FREngine on
certain types of processors

an error was returned on
IFRDocument::Process method call: IPE: .
\
src
\
AffinityM
askConvertor.cpp, 185.



HD #316762



Access violation error if the
Flush

method is called inaccurately.



HD #315853



If a large number of pages was saved with the
SaveTofile

method, an error was returned:
"Internal program error: .
\
src
\
DocumentPageImpl.cpp,

379 ".