Self-sufficient banti

Highlights

1. This system can process all image files by itself. Like jpg, png etc.
2. It can segment an image file by itself. It does need a box file from antanci.
3. It has its own segmentation code in page.py

antanci based system

This version need you to supply a box file as input. The box file is a pre-segmented image stored as glyphs.
So there is a huge dependency on the antanci ocr system. Later releases will not have this dependency.