Each ipynb file corresponds to a book, because the images scaned from different books have different noises.
The cropped results are in this folder: https://drive.google.com/drive/folders/1eDwHyiNaMLggn0giA-5hP7X61i0ppHbj
Please contact @tianrengao and @BohanZhai if you have any questions.
Here is an example from Banpo A Khrid Texts:
Before cropping the image:
After cropping, the image is split into 5 pieces:
image00005-0
image00005-1
image00005-2
image00005-3
image00005-4
