Show HN: Convert scanned documents into searchable PDFs

Post by jahewson » Wed, Dec 16, 2015, 11:48 PM UTC

jiaweihli wrote: post #15

cfcef wrote: post #9

Earlier quoted context omitted.

I hope not. Tesseract delivers bad results on high quality scans, far below the same OCR quality achieved by services like Google Books. What the OCR market needs is someone who will bring that level of OCR quality - or better - to the masses (perhaps some deep learning grad student with time to kill?), not yet another wrapper around Tesseract. We have those already!

Have you looked into ocropy[0]? Here's a nice intro[1] that later talks about how it achieves higher accuracy using an LSTM model[2]. [0] https://github.com/tmbdev/ocropy [1] http://www.danvk.org/2015/01/09/extracting-text-from-an-imag... [2] http://www.danvk.org/2015/01/11/training-an-ocropus-ocr-mode...

Note that the primary author of ocropy (formerly ocropus) works at Google.

Post by Cheyana » Thu, Dec 17, 2015, 1:14 AM UTC

bmh_ca wrote: post #4

While interesting, and looks to be a needed services, the page leaves many questions, such as: What's the privacy model? While the PDFs are deleted, what happens to the searchable content? Is it also deleted? What's the revenue model? How can we be sure it'll be around in a few months? Is there an AJAX interface? Is the quality or performance better than running Tesseract on a server?

Also relevant, how do we know they're not injecting a pdf exploit into the final document? There is no real company information to hold anyone accountable. There could be a dozen websites like this for people to use for "free" that will inject malicious script which most antivirus apps won't detect. Not saying it's not useful, and awesome if legit but there should be more accountability. This is almost the internet equivalent of a stranger in a car waving free candy at a child walking down the street. My workplace was hit yet again today with a Cryptolocker variant (second time this month) which required us to restore thousands of files from backup. All from clicking on a link in an email.

Re: Show HN: Convert scanned documents into searchable PDFs

Re: Show HN: Convert scanned documents into searchable PDFs