Re: [tesseract-ocr] Re: Store rotated pages

2017-07-07 Thread Thomas Klettke
Greg - here is the relevant part of the Perl code I've put together for my server, running CentOS 7. Feel free to modify for your use: == ## Search for pages that need to rotated # Arbitrary prefix for temporary files - # I'm also using the nanoseconds of

Re: [tesseract-ocr] Re: Store rotated pages

2017-07-07 Thread 'Greg Kinney' via tesseract-ocr
Hi Thomas, would you mind sharing your script? I've been desperately looking for a way to auto-rotate from the command line. On Tuesday, June 6, 2017 at 6:20:19 PM UTC-5, Thomas Klettke wrote: > > Thanks - I've figured it out, and have a solution that works now: > >- Input file is a

Re: [tesseract-ocr] Re: Store rotated pages

2017-06-06 Thread Thomas Klettke
Thanks - I've figured it out, and have a solution that works now: - Input file is a multi-page TIFF. - Some pages need to be rotated (90, 180, or 270 degrees - although other angles should work as well). - Output is a multi-page searchable PDF - Tools used: - Linux (Fedora

Re: [tesseract-ocr] Re: Store rotated pages

2017-06-01 Thread gmail
This didn't go through for some reason: " The free program imagemagic has many tool for manipulating image files. I don't use it very often but the last time I did, something like this worked (on command line): imconvert -rotate 91 *.jpg jrot-%04d.jpg I renamed the convert.exe binary to

Re: [tesseract-ocr] Re: Store rotated pages

2017-05-31 Thread gmail
The free program imagemagic has many tool for manipulating image files. I don't use it very often but the last time I did, something like this worked (on command line): imconvert -rotate 91 *.jpg jrot-%04d.jpg I renamed the convert.exe binary to imconvert.exe because windows will find a

[tesseract-ocr] Re: Store rotated pages

2017-05-16 Thread 'Michael Hecht' via tesseract-ocr
Ok, I did some basic analysis of the hocr file myself using python (analysing textangle and baseline) and then called imagemagic convert -rotate on the basic tif file. So finally I have an fully integrated workflow using convert, python and tesseract to convert a rotated, scanned pdf to a