In beginning of August 2010 I decided to maintain and to improve python py/tesseracttrainer.py script. Script was quit old, but with small improvements I was able to run it on Windows and Linux. I like very much its layout – how it display image and boxes.
While working with A5 scan with small number of boxes it works good. Bigger picture with a lot of boxes cases that application is not usable or did not start.
For this reason I was looking for something more efficient to system resources. I found tesseract-gui. But it was not developed actively for several month. After short discussion with author I choose to improve it under new name – QT Box Editor because of name conflict with other project. The target of project is simple: to provide efficient way for editing tesseract-ocr box file which is IMHO most important part of training new language.
It has different interface than pytesseracttrainer.py or cowboxer – similar to bbTesseract (last updated in July 2008). This is a good new for those who like to have more options ;-).
Version 1.02 is focused on zooming and icons. I use it on Linux and Windows without problem. I hope you find it useful.
Developers and contributors are welcomed.