For the scanned documents, I used to compressing them with jbig2 as one of the post-processing steps. Representing the pages as big images not only takes more

For the scanned documents, I used to compressing them with jbig2 as one of the post-processing steps. Representing the pages as big images not only takes more space to store/archive, but also increases IO-time for loading the documents.