
Source (link to git-repo or to original if based on someone elses unmodified work): Add the source-code for this project on opencode.net
KbookOCR — an intelligent system for recognition documents (OCR system).
Input: Specify document that you want to recognize (djvu, pdf, img ) + Select language of input document.
Pages to proceed: Scanning can be conducted of the entire document or selected range.
Preview size: Here is a few options of preview (located on the left preview window):
Native Small
Output: Outgoing documents can be saved in txt format (specify the folder you want to save) or opened with OpenOffice.
Based on: CuneiForm
Note: The quality of an output file depends of input source quality and work of third-party OCR package.
GitHub: http://github.com/b0noI/KBookOCR
PS: If you enjoyed our program, do not forget to click "+";) Or you can even donate
9 years ago
2.1 - better KDE integration, better UI, pre-build only for x32
2.0 - new major version of KBookOCR. All new:
- new GUI,
- new project system,
- new integration with cuneiform system,
- new scaner support system(KSane).
It's more stable, faster than 1.x version
1.4.1 "” you can load last project and continue to working on it
1.4.0 "” book pages thumbnails for recognition, batch scanning option
1.3.1: preview of scaned pages, some GUI usability improvements
1.3: new GUI
1.2: output in rtf, html (layout support)
scanner support (via scanimage)
GUI changed
UPD3: try to compile ONLY with Qt >= 4.7
UPD2: src is here, enjoy
UPD: rpm and binary tar.gz for all x32 dsit is here;) src will coming soon
9 years ago
2.1 - better KDE integration, better UI, pre-build only for x32
2.0 - new major version of KBookOCR. All new:
- new GUI,
- new project system,
- new integration with cuneiform system,
- new scaner support system(KSane).
It's more stable, faster than 1.x version
1.4.1 "” you can load last project and continue to working on it
1.4.0 "” book pages thumbnails for recognition, batch scanning option
1.3.1: preview of scaned pages, some GUI usability improvements
1.3: new GUI
1.2: output in rtf, html (layout support)
scanner support (via scanimage)
GUI changed
UPD3: try to compile ONLY with Qt >= 4.7
UPD2: src is here, enjoy
UPD: rpm and binary tar.gz for all x32 dsit is here;) src will coming soon
kolyacher
6 months ago
Report
janopi
4 years ago
Thank you
(sorry by my english)
Report
SilentSpirit
6 years ago
Report
MatthiasM
9 years ago
https://build.opensuse.org/package/live_build_log?arch=i586&package=kbookocr&project=home%3AMailaender&repository=openSUSE_12.1
/usr/lib/gcc/i586-suse-linux/4.6/../../../../i586-suse-linux/bin/ld: kbookocr.o: undefined reference to symbol 'KIcon::~KIcon()'
/usr/lib/gcc/i586-suse-linux/4.6/../../../../i586-suse-linux/bin/ld: note: 'KIcon::~KIcon()' is defined in DSO /usr/lib/libkdeui.so.5 so try adding it to the linker command line
/usr/lib/libkdeui.so.5: could not read symbols: Invalid operation
collect2: ld returned 1 exit status
make: *** [KBookocr] Error 1
error: Bad exit status from /var/tmp/rpm-tmp.eLtV7J (%build)
Report
sosonok
9 years ago
Today I've 'ocred' a 104 pages pdf. Result is decent, but there are a few issues:
-even if I select document language, some characters are not recognized (—, not -)
-every line that ends with '-' splitting a word results in a broken paragraph.
-output has no format (bold, alignment, font size, margins).
Can this be fixed or is a cuneiform limitation?
Report
gohanz
9 years ago
viewadder.cpp:33:6: warning: unused parameter ‘doc’
viewadder.cpp:33:6: warning: unused parameter ‘n’
g++ -c -O2 -march=i486 -mtune=i686 -O2 -march=i486 -mtune=i686 -Wall -W -D_REENTRANT -DQT_NO_DEBUG -DQT_GUI_LIB -DQT_CORE_LIB -DQT_SHARED -I/usr/lib/qt/mkspecs/linux-g++ -I. -I/usr/lib/qt/include/QtCore -I/usr/lib/qt/include/QtGui -I/usr/lib/qt/include -I/usr/include/poppler/qt4 -I. -I. -o scanerdialog.o scanerdialog.cpp
scanerdialog.cpp:2:29: fatal error: ui_scanerdialog.h: File o directory non esistente
compilation terminated.
make: *** [scanerdialog.o] Errore 1
make: *** Attesa dei processi non terminati....
ocrthread.cpp: In member function ‘bool OCRThread::startOCR()’:
ocrthread.cpp:168:1: warning: no return statement in function returning non-void
ocrthread.cpp: In member function ‘QString OCRThread::getImgAt(int)’:
ocrthread.cpp:238:1: warning: control reaches end of non-void function
Report
414N
9 years ago
Report
b0noI
9 years ago
Report
gohanz
9 years ago
Report
olegue
9 years ago
2. How about OCR-ing multilingual languages documents ?
3. How to OCR-ing a part of the image ?
Report
b0noI
9 years ago
Report
olegue
9 years ago
The Layout option is a good move but needs polishing. It arranges the output in different order then the original.
Report
doktor5000
9 years ago
first i'd like to thank you for your efforts. When building KBookOCR for Mageia 1 x86_64 i noticed that there is a library hardcoded with path. See for yourself:
[[email protected] KBookocr]$ grep -R libksane.so ./
./KBookocr.pro:/usr/lib/libksane.so
./Makefile:LIBS = $(SUBLIBS) -L/usr/lib64 -L/usr/lib -lpoppler-qt4 /usr/lib/libksane.so -lQtGui -L/usr/lib64 -lQtCore -lpthread
This is no good and breaks build on x86_64. Please fix with next release.
Report
b0noI
9 years ago
Report
mityukov
9 years ago
Report
b0noI
9 years ago
Next we plan to support multiple drivers (not just cuneiform). There will also be implemented many new features including for example: automatic language detection, improve the recognition of document structure, etc...
KBookOCR finally killed all analog Books)
Report
Martox12
9 years ago
I am running Kubuntu 11.04 amd64. When installing package 'kbookocr_2.0.5.x64.deb' using 'gdebi-gtk' from the konsole ... everything goes OK until the end ... when I receive this screenshot ( http://farm6.static.flickr.com/5067/5879629016_f367c0f77c_z.jpg ). The terminal says nothing. But I can run the program, and things appear to go nice so it's a bit strange. I have been digging a bit and I don't know if this is ok (a screenshot from synaptic) ( http://farm7.static.flickr.com/6001/5879629020_6fde1285e7_b.jpg ) ... maybe it says something to you ... maybe this is not the problem.
On the other hand, I have found that cuneiform is in version '1.1.0+dfsg-1' at 'Oneiric Ocelot' ... and I was wondering if KBookOCR is ready for this version.
Congratulations for the application, it looks very promising !!!
Regards.
Report
b0noI
9 years ago
Report
Martox12
9 years ago
I am running Kubuntu 11.04 amd64. When installing package 'kbookocr_2.0.5.x64.deb' using 'gdebi-gtk' from the konsole ... everything goes OK until the end ... when I receive this screenshot ( http://farm6.static.flickr.com/5067/5879629016_f367c0f77c_z.jpg ). The terminal says nothing. But I can run the program, and things appear to run nice so it's a bit strange. I have been digging a bit and I don't know if this is ok (a screenshot from synaptic) ( http://farm7.static.flickr.com/6001/5879629020_6fde1285e7_b.jpg ) ... maybe it says something to you ... maybe this is not the problem.
On the other hand, I have found that cuneiform is in version '1.1.0+dfsg-1' at 'Oneiric Ocelot' ... and I was wondering if KBookOCR is ready for this version.
Congratulations for the application, it looks very promising !!!
Regards.
Report
amoskvin
9 years ago
Report
b0noI
9 years ago
Report
maidis
9 years ago
$ sudo pisi it -c system.devel
$ sudo pisi bi https://svn.pardus.org.tr/pardus/playground/maidis/ocr/cuneiform/pspec.xml
$ sudo pisi it cuneiform*.pisi
$ sudo pisi bi https://svn.pardus.org.tr/pardus/playground/maidis/ocr/kbookocr/pspec.xml
$ sudo pisi it kbookocr*.pisi
I made also a desktop file [2] for adding a entry to KDE menu and a patch [3] for fixing compilation on Pardus 2011. Could you add these to KBookOCR, if theye are OK?
Do you plan to support other OCR systems (Tesseract, GOCR, Ocrad...)?
[0] https://svn.pardus.org.tr/pardus/playground/maidis/ocr/kbookocr/
[1] http://www.pardus.org.tr/eng/
[2] https://svn.pardus.org.tr/pardus/playground/maidis/ocr/kbookocr/files/kbookocr.desktop
[3] https://svn.pardus.org.tr/pardus/playground/maidis/ocr/kbookocr/files/add-kde4-include-dir.diff
Report
b0noI
9 years ago
Yes, we plan to support other engines for OCR.
For desktop file, it is present in the deb and rpm packages, but you certainly are right and I will add it in the src as well.
Report
sosonok
9 years ago
Report
b0noI
9 years ago
Report