Beyond transcription : Case studies in special document analysis requirements

Through case studies of two current projects at Johns Hopkins University’s Digital Knowledge Center (DKC), this paper discusses some document analysis applications other than straightforward transcription that were developed through direct communication with users of digital collections. These applications include lyrics extraction from sheet music, an image-based annotation collaboratory environment and automatic illumination-finding in medieval manuscripts. Some of the problems encountered developing new technology across a disciplinary divide are then discussed. This paper aims to foster discussion related to how to make document image analysis research more user-centric.