Recognition in maps and geographic documents: Features and approach

Abstract

The objective in the recognition of maps and geographic documents is to extract the topological configuration among composite elements (or objects) or to distinguish the constructive relationships among related composite elements as well as to recognize composite elements individually. Of course, the maps and geographic documents reflect the structural features mapped deformedly from the corresponding geographic configuration in the real world. Some composite elements are often overlayed and intersected mutually. This drawing feature makes it difficult to accomplish the recognition objective smartly. In this paper, we discuss the application-dependent features, approaches and paradigms in the recognition of maps and geographic documents. In particular, we intend to make the recognition framework of maps and geographic documents clear so as to be distinguished characteristically for graphics recognition or drawings interpretation with respect to the above discussion viewpoints.