Optical Character Recognition engine with stroke width transform(swt), maximally stable extremal regions, connvolutional neural networks, and various morphological operations
Dependencies required are
numpy==1.14.3, opencv-python==3.4.0.12, Pillow==5.1.0, pypillowfight==0.2.4, autocorrect, spellchecker, editdistance, tensorflow,
Usage
To run the OCR engine, use following command
python3 main.py demo.png demoGroundTruthText.txt
After running this, you'll see result.txt, with extracted texts, and coordinates and width&height of the text region, you'll also get textBlockdemo.png which is an image with bounding box indicating which part has been extracted. To compare with pytesseract using:
python3 comparePytesseract.py demo.png demoGroundTruthText.txt
MISC
To get a better understanding, have a look in the report
We use 'char74k' to train the CNN model.