|
OCR: Destroying the Text In Order to Save It OCR doesn't work. It sort of works, and it works fine if you are trying read well-printed text in Courier or Helvetica. It doesn't work on anything printed before 1925, or faded, or with broken type, or with Italics, or with footnotes, or with paper that has turned brown. You might as well type the text in yourself. My pages from the anthropological society of Bombay will be scanned in and preserved as an image, or typed in by hand. OCR won't do it. Some day this problem will be solved, and some day I will be able to start up a process and the text will automagically appear on my web site. We should all live that long.
|