Dear all,
I would like to evaluate the OCR output quality according to preprocessing made on documents.
I found the script editdist.lua which may help. It works and gives results but :
- I don't know what the numbers refer to.
- Is it possible to get the error rate in % ?
Any help ?
Thank!
Guillaume