0
I would like to scan old text documents, and then destroy some of the originals. There would be a few thousand sides of A4.
To do this I will need some procedure to "QC" that all the pages have been scanned, and that the scanned pages are readable.
Is there any way to do this other than spot checks?
I'm not aware of a reliable method to automatically assess scan quality for a critical application. The thing is, you can glance at an image of a page and know if there are problems. It would be worth the time to simply look at the results yourself. It would only take a second or two per page. – fixer1234 – 2015-03-29T18:48:57.910
I think you're right that a quick check of each page is practical and wise. But I'm not sure that will catch missing pages, nor small areas of text that are not legible. Perhaps small bad areas is sufficiently rare that it's not a problem? Surely that won't be true of missed pages though? I guess the best way to catch those after the batch is fed through is to examine each physical page before inspecting its scan and put it into a "done" pile after checking the scan exists. – Croad Langshan – 2015-03-29T19:47:48.917
There is no way to scan a page twice, so a page count for the batch will tell you if a page was missed. Scan problems are surprisingly easy to spot. On a good quality original, any defect will stand out. The pages with scan problems will be ones with problems on the original (physical condition or content quality), and those pages will be easily recognized. Just pay closer attention to those. If a high percentage of your originals are not in good physical condition, or the content has problems (like low contrast, discoloration, etc.), it might require a different kind of scanning process. – fixer1234 – 2015-03-29T20:30:16.590