Lewisc
Member
I've been having issues where the OCR is not reading the characters correctly, for example Notting is being displayed as Netting, this is affecting the results that our users are having and influencing our internal research, is there anything that can be done?
Thanks
admin
Key Master
OCR is dependent on a number of factors.
The key issue is to ensure that you are scanning at the right resolution. Ideally, you want to be scanning at 300dpi. You could use a higher resolution, but for the purposes of R/KYV, 300dpi is the optimum, allowing for reasonable file sizes, and good quality OCR.
Also, the quality of the underlying document is important. The more degraded the paper copy, the worse the OCR.
While R/KYV does de-skew images, again, the better the quality of the original image, the better the OCR results.
If all of those factors are satisified, though, you might want to consider an upgrade to the OCR. R/KYV's OCR is now over 6 years old, and there have been significant advances since then in OCR technology.
admin
Key Master
OCR is dependent on a number of factors.
The key issue is to ensure that you are scanning at the right resolution. Ideally, you want to be scanning at 300dpi. You could use a higher resolution, but for the purposes of R/KYV, 300dpi is the optimum, allowing for reasonable file sizes, and good quality OCR.
Also, the quality of the underlying document is important. The more degraded the paper copy, the worse the OCR.
While R/KYV does de-skew images, again, the better the quality of the original image, the better the OCR results.
If all of those factors are satisified, though, you might want to consider an upgrade to the OCR. R/KYV's OCR is now over 6 years old, and there have been significant advances since then in OCR technology.
You must log in to post.