We want to develop a platform: web/mobile which can recognize Arabic law documents and do the search on them.
The main concern we have now related to the accuracy of the cloud OCR. - Can we improve the accuracy rate if we integrate your SDK on our servers? - Can we somehow detect/highlight the badly recognized words for further manual correction?
Looking forward to your soon reply.
Best regards, Ilya
asked 15 Nov '16, 09:47