@article{oai:nagoya.repo.nii.ac.jp:00005343, author = {Watanabe, Toyohide and Zhang, Rui}, journal = {Proceedings of the Fourth International Conference on Document Analysis and Recognition}, month = {Aug}, note = {The map recognition is one of interesting subjects to distinguish the meaningful information automatically from map images and construct resource data in information systems such as GIS (Geographic Information System). However, it is difficult to identify individual information because the map components are mutually interested or overlayed, and because the properties of map components are not well defined. In this paper, we propose an extraction method of character strings with map composition rules. Our recognition process is composed of two phases: extraction of character strings; and classification of character strings. The extraction phase is organized by a bottom-up approach, based on the measurement/estimation among pixels, while the classification phase is composed by means of a top-down approach, based on the interpretation/validation among map components.}, pages = {805--808}, title = {Recognition of character strings from color urban map images on the basis of validation mechanism}, volume = {2}, year = {1997} }