Hi all,

I’m not blogging a lot since i produce weekly report on gnome-soc-list. Also, the motivation is not that high. I continue the new preview area system. It is quite nice. It use GtkStyle colors and is extensible. That’s not easy to rewrite such software. In some hand you got excitation and motivation due to new features and huge improvement, but in the other end, its a bit of duplication, waste of time, etc.

After preview area, the big job is processing. I prepared a bit that part of the software, but wasn’t able to know exactly what i wanted. I guess i will provide a basic processor for rotation and color correction, however i don’t know how to get it really modular. Should i use multiple processors ? Also, i wonder how to take hardware color correction in account as well as calibration.

Anway, i would like to write OCR as soon as possible. Writing a dumb plugin for gedit or abiword would be funny. I tested OCRopus. Well, tesseract SVN did not build so ocropus can’t be built. tesseract development seems not active. I know some people are working on it, but not in the official SVN. Finally, i found a patch in their bugtracker which solved the problem. Tesseract and OCRopus definitly needs some feedbacks. This is where gnome-scan comes at point.

Even if OCRopus is far from yet ready for distribution, i investigate into it. Gnome Scan en Gegl are is similar state. I wish this will help and motivate their developers. Some projects really needs some love. Let’s give it what they need.

OCRopus should require an additional dialog with preview of the result. I bet the new preview area will be suitable for this thanks to it’s modular design. The new preview area will also fit exotic needs like multiple area without bloating the library.


3 Responses to “Finishing the rewrite ; new issues ; toward processing”

  1. gicmo Says:

    Did I get it right that you would do OCR stuff already in the preview area?
    I am doing a lot of book scans for my studies and the workflow of 1) scanning the pages 2) cutting all the pages and then 3) do the ocr (including correction of all the maybe worningly recognized words) and finally in the last step make a pdf out of it which has the picture as presentation for the pages and the text behind it so you can select and search.

    I am not sure that should all happen in the scan dialog since scanning the image and then post-processing is a totally different task for me (since I might do some more/better ocr on the saved pictures later anyway) ..

    But maybe I did get you totally wring ..

  2. Étienne Bersac Says:

    Hi Gicmo,

    By reusing preview area, i mean reusing the class in a third dialog, not the instance of the first dialog ;).


  3. Alexandre Says:

    Welcome to OpenIcc then :)

Comments are closed.