Paragraphs instead of lines in gpageseg

[xdsv]

Is there a way to get paragraphs instead of lines in ocropy-gpageseg?
If that's not possible, from what I saw only the .pseg file contains the paragraph information encoded as a different color. Is there a utility to extract that information?

[zuphilip]

No, the page segmentation will always output lines. However during this process it tries to identify the columns (I am not sure that also paragraphs are identified in this step) and this information is saved in the pseg files, as you noted already. See in the wiki for more information about the file format.

Maybe try also to go further and create a hocr file with ocropus-hocr which should contain some paragraphs splits as <p />.

[xdsv]

Thanks for the reply, actually what I'm looking for is find blocks of text in an image. Can ocropy do this?

…

On Mon, Jun 26, 2017 at 8:23 AM, Philipp Zumstein ***@***.***> wrote: No, the page segmentation will always output lines. However during this process it tries to identify the columns (I am not sure that also paragraphs are identified in this step) and this information is saved in the pseg files, as you noted already. See in the wiki for more information about the file format <https://github.com/tmbdev/ocropy/wiki/OCRopus-File-Formats#physical-layout> . Maybe try also to go further and create a hocr file with ocropus-hocr which should contain some paragraphs splits as <p />. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#228 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AcUdDkJZfJsDCtwyCA59J1vR-fRHfSbjks5sH0BYgaJpZM4OEve-> .

[Sultan91]

Also interested in this question. In addition, it is not quite clear how text line ordering works