Python program to classify persian characters, including letters and digits.
- Gathered datasets of handwritten persian.
- Used the Lenet architecture to design two models: letter classifier and digit classifier.
- Feeded the neural networks with form images, filled with information.
- Used OpenCV function to recognize which check box of the input image has been filled.
- Achieved a test accuracy higher than 90%.
Giving the image forms like below, this program is able ro recognize the firstname, lastname, student number and degree.
