Document analysis systems V

This volume contains papers selected for presentation at the5thIAPRWo- shoponDocumentAnalysisSystems(DAS2002) in Princeton, NJ on August 19–21, 2002. As such, it represents the contributions of an international com- nity of academic, industrial, and government researchers and re?ects the state of the art in such diverse domains as character recognition and classi?er design, text segmentation, table understanding, page layout analysis, and document - gineering, indexing, and retrieval, along with emerging application areas incl- ing the World-Wide Web, where algorithms developed for “traditional” paper documents take their place alongside completely new techniques. DAS 2002 continues in the ?ne tradition of past workshops held in Kaise- lautern,Germany(1994),Malvern,PA(1996),Nagano,Japan(1998),andRiode Janeiro, Brazil (2000). DAS is distinguished from other gatherings of document analysis researchers by its emphasis on systems. To extract useful information from a noisy page image is a complex process requiring multiple stages, each of which is a research topic in its own right. The integration of such components introduces additional complications. Hence, building and testing document a- lysis systems is a challenging task for which no de?nitive solution yet exists. You will ?nd that the papers in this volume carry with them a “?avor” that is unique to DAS. The 44 regular and 14 short papers that appear herein derive from 15 di- rent countries, highlighting the international make-up of our discipline. Regular papersweresubjectedtorigorousreviewbymembersoftheProgramCommittee as well as other noted researchers. Short papers were chosen by the Co-Chairs for their relevance to the themes of the workshop.