ocr

Text recognition from an images

Create Dataset

create dataset with structure:

    dataset
    |
    ...train
    ...|
    ......classA
    ......|
    .........imgA
    .........imgB
    ............
    .........imgZ
    ......classB
    ...
    ...validation
    ...|
    ......classA
    ......classB
    ...

Set CLASSES in ./main.py Change Dense with class count in ./train.py near model = Sequential([