Curious.
I've never used the OCR results from Google. I download the PDF file, convert it to images, clean it up with ScanTailor and the use Abbyy Sprint to do the OCR. Sprint came with my scanner. The conversion can be done with PDFill if you're using Windows or pdftopng if using Windows or Linux and you are comfortable with command line. pdftopng is faster. This how I've done the OCR for the Markham poetry books which continue to proceed at a very desultory manner. By the end of the year as well as the rest of the Thorne Smith books. Soon... Real soon now.
|