MobileRead Forums - View Single Post

brecklundin · 09-25-2010, 01:24 AM

if you have a DSLR or can borrow one when needed maybe a nice used copy stand with a nice flat-field lense mounted to your camera is a better option and faster. You can shoot RAW, batch adjust the white balance, sharpen and correct any perspective issues then batch convert to TIFF for your OCR software to grind on...then the fun begins in cleaning up the text...and the issue of graphic content in your images. Or you can just create a set of images and print to a PDF which won't be searchable but will maybe be enough.

BTW, flat field lenses are designed for shooting 2D stuff like books, paintings and such and are pretty much distortion free in their design...the copy stand holds the camera and lens in the same position and you just put the book or pages you are copying on the tray of the stand which is made so you put everything in the same position so after the first page is setup it's just a boat race to get it done...it's a very, very fast process for books, especially when compared to consumer grade scanners. You can even get by with a point and shoot digital camera is it has a good macro mode, some can even shoot RAW or TIFF natively. I would try and use DNG format for RAW not because it's better but more software reads that format compared to some RAW formats out there.

While I would use Lightroom to do the file adjusting and editing in batches in preparation for the OCR conversion, Photoshop can create PDF files in case you did not know. I've only tried it a few times and obviously did not really have much of a clue if you look at the results, but it worked well enough for my needs.

I just mention this to point out using a slow scanner is not the only, or maybe not always the best option/method...

09-25-2010, 01:24 AM	#5
brecklundin Banned Posts: 1,906 Karma: 15348 Join Date: Jun 2007 Device: mine	if you have a DSLR or can borrow one when needed maybe a nice used copy stand with a nice flat-field lense mounted to your camera is a better option and faster. You can shoot RAW, batch adjust the white balance, sharpen and correct any perspective issues then batch convert to TIFF for your OCR software to grind on...then the fun begins in cleaning up the text...and the issue of graphic content in your images. Or you can just create a set of images and print to a PDF which won't be searchable but will maybe be enough. BTW, flat field lenses are designed for shooting 2D stuff like books, paintings and such and are pretty much distortion free in their design...the copy stand holds the camera and lens in the same position and you just put the book or pages you are copying on the tray of the stand which is made so you put everything in the same position so after the first page is setup it's just a boat race to get it done...it's a very, very fast process for books, especially when compared to consumer grade scanners. You can even get by with a point and shoot digital camera is it has a good macro mode, some can even shoot RAW or TIFF natively. I would try and use DNG format for RAW not because it's better but more software reads that format compared to some RAW formats out there. While I would use Lightroom to do the file adjusting and editing in batches in preparation for the OCR conversion, Photoshop can create PDF files in case you did not know. I've only tried it a few times and obviously did not really have much of a clue if you look at the results, but it worked well enough for my needs. I just mention this to point out using a slow scanner is not the only, or maybe not always the best option/method...