jTessBoxEditor

[quân]

jTessBoxEditor is a box editor for Tesseract OCR data, providing editing of box data of both Tesseract 2.0x and 3.0x formats. It can read images of common image formats, including multi-page TIFF. The program requires Java Runtime Environment 6.0 or later.

http://vietocr.sourceforge.net/training.html

[quân]

Version 0.6 adds a utility function that creates TIFF/Box pair suitable for training with Tesseract. The box generated is nearly identical to the one generated by tesseract; it can be used to validate one another. You can use a Unicode-compatible file compare tool, such as WinMerge, to compare the box files.

[quân]

jTessBoxEditor v0.7 has been released with the following enhancements:

- Fix an issue with opening Help file on OS X
- For TIFF/Box generation:

- increase line spacing
- abbreviate bold/italic font style to b/i for filename
- add a Prefix (Language Code) textbox
- add support for text anti-aliasing

Also, the PowerShell script train.ps1 has been updated to automate training with Tesseract 3.02 on Windows platform.

[quân]

jTessBoxEditor v0.8 has been released with the following enhancements:

- Add row number header
- Char cell now editable
- Convert Unicode escape sequences where possible
- Find box now displays Unicode characters and allows search using Unicode escape sequences
- Improve Generate TIFF/Box functionality:

* automatically combine boxes that have the same coordinates or completely encloses one another
* automatically combine boxes that are combining symbols, specified in an external file, with the main, base character
* retain last-modified exp number in filename

http://vietocr.sourceforge.net/training.html

[quân]

jTessBoxEditor v0.9 Release:

- Enhance Generate TIFF/Box functionality to allow for combining prepending symbols in addition to appending
- Fix a bug that failed to persist changes to table in edit mode
- Find function now supports partial matches
- Fix a problem with table not scrolling along when row header has focus and scrolling