I acquired an old Canon Canoscan N650U USB scanner recently. These date from the turn of the century, so I plugged it into my Ubuntu 10.10 system “on the off chance it might still work”, and to my surprise it (a) was immediately recognised by Ubuntu, and (b) still works perfectly.
The “Simple Scan” application included with Ubuntu is limited to the essential options, and is none the worse for that. You can crop the scan area, change the resolution, and select between text and image scan types, but that’s it. The resulting scan can be saved as one of JPEG’s, PNG’s (which can be rather large) or PDF’s.
Maximum supported resolution at the application level is 2400dpi, which is about half the marketing claims for current generation USB scanners, but oddly a lot more than the 600x1200dpi optical resolution claimed by the 650U’s specifications. I’m guessing the hardware does some interesting interpolation to produce the higher resolution output, which makes me wonder if best results might be obtained by setting it to 600dpi or 1200dpi? Either way, the results seem more than adequate for my needs so far. It’s certainly looks “good enough” from the scans of text documents and photo-type images that I’ve done so far, and ought to let me do simple one-off “photocopying” if I need to.
What I’d like to do next is find a simple OCR application that can work in tandem with the simple scan application to produce plain text output. There seem to be several OCR offerings that work under linux, but it’s clear that none of them are integrated into the basic system in the same way that the scanner support is.
A little more investigation is required.