http://www.danielstender.com/granthinam/564/ Daniel Stender's blog on Sanskrit philology and associated issues (e-philology, textual criticism, TeX, Open source, etc.) Wed, 28 Jul 2010 11:42:51 +0000 hourly 1 http://www.danielstender.com/granthinam/564/comment-page-1/#comment-1236 Daniel Stender Mon, 21 Jun 2010 09:25:59 +0000 http://www.danielstender.com/granthinam/?p=564#comment-1236 Update: (1) I really haven't thought about kicking off such a thing with this but the American <em>Linux Magazine</em> has bought the article from German <em>Linux User</em> which has emerged from this contrib (issue 06/2010), and it's going to appear in no. 117 in August. (2) In the meanwhile Abbyy published a Linux command line version of their popular <em>FineReader</em> software (see <a href="http://translate.google.de/translate?js=y&prev=_t&hl=de&ie=UTF-8&layout=1&eotf=1&u=http%3A%2F%2Fwww.pro-linux.de%2Fnews%2F1%2F15418%2Fabbyy-veroeffentlicht-command-line-interface-ocr-fuer-linux.html&sl=de&tl=en" rel="nofollow">here</a>), which seems to work really fine (see the review in the recent <em>Linux-Magazin</em>, 07/2010, 68-71). The reader spits out also hOCR but also finished PDFs. (3) There is a comparison of the free OCR engines in the <em>Linux-Magazin</em> 12/2006 (see <a href="http://translate.google.de/translate?js=y&prev=_t&hl=de&ie=UTF-8&layout=1&eotf=1&u=http%3A%2F%2Fwww.linux-magazin.de%2Flayout%2Fset%2Fprint%2Fcontent%2Fview%2Ffull%2F487&sl=de&tl=en" rel="nofollow">here</a>, might be a little bit outdated). (4) Jakub Wilk also implemented a little GUI editor for DjVu metadata, <a href="http://jwilk.net/software/djvusmooth" rel="nofollow">Djvusmooth</a>, which is available as I can see so far only for Debian Testing (that's anyway the choice if you want to run the latest stuff), see <a href="http://www.telemedia.ch/cgi-bin/whohas.wh?dist=alldists&q=djvusmooth&go=Search!" rel="nofollow">here</a>. Update:
(1) I really haven’t thought about kicking off such a thing with this but the American Linux Magazine has bought the article from German Linux User which has emerged from this contrib (issue 06/2010), and it’s going to appear in no. 117 in August.
(2) In the meanwhile Abbyy published a Linux command line version of their popular FineReader software (see here), which seems to work really fine (see the review in the recent Linux-Magazin, 07/2010, 68-71). The reader spits out also hOCR but also finished PDFs.
(3) There is a comparison of the free OCR engines in the Linux-Magazin 12/2006 (see here, might be a little bit outdated).
(4) Jakub Wilk also implemented a little GUI editor for DjVu metadata, Djvusmooth, which is available as I can see so far only for Debian Testing (that’s anyway the choice if you want to run the latest stuff), see here.

]]> http://www.danielstender.com/granthinam/564/comment-page-1/#comment-1053 Daniel Stender Tue, 27 Apr 2010 08:07:46 +0000 http://www.danielstender.com/granthinam/?p=564#comment-1053 Thanks Edgard. Indeed everything here is under heavy development (actually Debian [Testing] is the best distrib for running all the newest stuff). I've been in contact with Jakub Wilk and he told me grayscale processing will be also available soon (with 0.4.2, option <em>-render</em>). BTW for proper single word highlighting <em>hocr2djvused</em> needs data with has been produced with <em>ocroscript --charboxes</em>! Thanks Edgard. Indeed everything here is under heavy development (actually Debian [Testing] is the best distrib for running all the newest stuff). I’ve been in contact with Jakub Wilk and he told me grayscale processing will be also available soon (with 0.4.2, option -render). BTW for proper single word highlighting hocr2djvused needs data with has been produced with ocroscript –charboxes!

]]> http://www.danielstender.com/granthinam/564/comment-page-1/#comment-1051 Edgard Bikelis Tue, 27 Apr 2010 03:51:53 +0000 http://www.danielstender.com/granthinam/?p=564#comment-1051 Excellent post, Daniel. The Ocrodjvu is really a good finding, as the last time I've checked, we could just add OCR to each line, not to each word. Now, being able to search all my .djvu files will be neat indeed! Excellent post, Daniel. The Ocrodjvu is really a good finding, as the last time I’ve checked, we could just add OCR to each line, not to each word. Now, being able to search all my .djvu files will be neat indeed!

]]> http://www.danielstender.com/granthinam/564/comment-page-1/#comment-559 Daniel Stender Tue, 02 Mar 2010 15:37:06 +0000 http://www.danielstender.com/granthinam/?p=564#comment-559 Latest Gscan2pdf 0.9.30 employs a port to OCRopus, cf. http://lwn.net/Articles/372405/ Latest Gscan2pdf 0.9.30 employs a port to OCRopus, cf. http://lwn.net/Articles/372405/

]]> http://www.danielstender.com/granthinam/564/comment-page-1/#comment-511 Harunaga Isaacson Mon, 08 Feb 2010 11:50:48 +0000 http://www.danielstender.com/granthinam/?p=564#comment-511 Thanks for the post. gscan2pdf is indeed very nice; I used to use it a lot. Have you tried Scan Tailor (http://scantailor.sourceforge.net/ )? I have found this useful too. Thanks for the post. gscan2pdf is indeed very nice; I used to use it a lot. Have you tried Scan Tailor (http://scantailor.sourceforge.net/ )? I have found this useful too.

]]>