How difficult is it to recognize optical character recognition (OCR) formally? Assume that the margin of error is comparable to that of a person (which, in my opinion, is about 98%).
In other words, where would it correspond to the complexity scale P / NP complexity and intractability?
Or will it fit that scale? What is the problem?
I am not very familiar with the formal definition of the complexity of a problem. I'm just curious.
source
share