Skip to Main Content
The problem of hiding data in a document image is considered. Drawbacks of existing techniques, particularly when applied to documents written in non-English languages, are addressed. A new data hiding technique for document image is proposed. The new method offers high data embedding capacity, and is applicable to a wider class of languages than the existing ones. Results from experiments using both simulated data and real photocopied document are provided to demonstrate practical usefulness of the proposed method.