Numeric Recognition Program Project
- Get camera data.
- Remove as much external background as possible through a filter.
- Recognize through OCR.
This is what we are doing now. However, now it is a form that works well only in one environment.
What method would be good to recognize this more accurately and in detail? Iโm curious if you have any good ideas or techniques.
help me.
Iโm curious about your opinion.