README.md

Handwriting Recognition

Arda Mavi - ardamavi.com

Handwriting recognition with machine learning.

I used decision trees from Scikit-Learn to finding which character in image. Scikit-Learn is an open source Python library for machine learning.

Using Commend:

python3 main.py <ImageFileName>

Adding new characters and train dataset:

If you want to add new character to datasets, you create a directory and rename what you want to add character (like 'a' or 'i').

If you want to add a new training images to previously character datasets or null dataset of character, you add a image(suggested image size: 25x25) to about character directory.

Note: We work on grayscale(2D matrix) image also if you use color images(like RGB), program will automatically return to grayscale from your color image.

Used Modules:

• sys
• numpy
• PIL (Image)
• os (listdir)
• sklearn (joblib and MLPClassifier)

Important Notes:

• Suggested image size: 25x25
• Install above modules
• If you want to update classifier(maybe for newly trained classifier with new data sets), delete 'Data/classifier/classifier.pkl' file and run program again.
• The program can only distinguish '0', '1' and 'A' characters. To be able to distinguish other characters, you should add new character datasets.