The new interesting stuff in this version is that OCRFeeder can now export to PDF. When exporting the pages to PDF, users will have two choices: “a PDF from scratch” or “a searchable PDF”. The PDF from scratch means that the text part of what will be exported will be written in the PDF using ReportLab whereas the searchable PDF means that the PDF will present the whole original picture but with invisible text overlaid in order to make it searchable.
The PDF exportation still needs some polishing but I wanted to get it out there as soon as possible for the people who need it.
Check out these examples:
(page loaded in OCRFeeder and recognized automatically)
(exported PDF from scratch)
(exported searchable PDF with selected text)
This version also fixes issues when recognizing grayscale pictures as well as the mouse cursor that was being changed when it was over a page’s right margin.
I’ve also added separators to divide the Document’s submenus so they are grouped correctly and I’ve made ODT the first choice in the list of exportation formats, which had been mistakenly changed.
As usual, the incredible team of translators is doing a great job and apart from the updated translations, OCRFeeder now comes in Catalan (with the Valencian option as well) and in Greek.
No, once again, OCRFeeder’s talk wasn’t approved by the Desktop Summit’s organization. If you think that I’ve presented it some well known conferences (LinuxTag, GUADEC ES and twice in FOSDEM), it makes me a bit sad that I couldn’t yet present this unique project in the conference of the desktop it targets, but let’s hope it makes it next year.
Still, Igalia is sponsoring me again to attend the DesktopSummit, so, if you’re interested in OCRFeeder or other projects I’m involved, let me know!
This version has no big improvements and exists mainly to introduce the fix of a bug that prevented using the algorithm for recognizing documents automatically.
The copyright was updated to include the proper copyright and license notices of ODFPy, which ships with OCRFeeder.
It also features some improvements to Debian related files (thanks to Alberto Garcia, who is creating the official deb package for Debian) and a few translation updates.
* Importing PDF files is now faster
* The OCR engines manager dialog now allows to detect and choose to use system-wide OCR engines (this action is also used when the application is started with no engines configured)
* Multiple content areas in OCRFeeder’s canvas can now be selected using Shift+Click
* Introduces Ctrl+a shortcut to select all content areas in OCRFeeder’s canvas
* The Tools menu now has the new action “Recognize Selected Areas” which will perform the automatic recognition on selected content areas of OCRFeeder’s canvas
Also, a few bugs were fixed:
* Removed PDF files’ extension from the images generated from them
* Sorts images when adding them from a folder
* Selection areas are now getting selected after creating them
* Fixed problem when quitting the application