Closed cathal-c closed 7 years ago
So the OCR works a treat. Once I trained the specific font I needed, it identifies characters with extreme accuracy.
Unfortunately however, in my case there exists symbols between certain characters (German license plates).
Looking at the debug console, these symbols are all well below 60% confidence, however they are outputted in the string regardless.
Is there anyway to easily modify the required confidence of the OCR for a character to be added to the output?
For anyone else who wishes to modify the confidence level, it can be located in SwiftOCR.swift on line 57.
So the OCR works a treat. Once I trained the specific font I needed, it identifies characters with extreme accuracy.
Unfortunately however, in my case there exists symbols between certain characters (German license plates).
Looking at the debug console, these symbols are all well below 60% confidence, however they are outputted in the string regardless.
Is there anyway to easily modify the required confidence of the OCR for a character to be added to the output?