Document image analysis (DIA) and digital libraries (DL) are working together to develop new technologies capable of transforming paper documents from the world heritage into reusable electronic documents. This new trend is well explained by the increasing number of digitization projects in the world, which sustain an important research effort tackling the problem of automatically processing the growing mass of digitized documents. The challenges of both research fields, document image analysis and digital libraries, concern the diversity of the document collections and the various metadata, which need to be extracted to satisfy the users needs, as well as the requirements of the libraries and the limits of the current technology.