Study Of Topographic Maps To Identify And Road Network Data Model

This paper mainly presents the way to map symbol recognition based on decision tree. Through research on the structure design and discrimination decision making about decision tree, recognition process of map symbol based on decision tree learning is provided. The research emphasizes on the feature selection in internal-nodes, and designs a new Distance Difference Function. DDF solves the intrinsic bias of information-gain to some extent, which enhances the robustness of recognition algorithm. Meanwhile, DDF helps to form the tree structure with proper depth and width, which improves the efficiency of tree search.In addition, Road data model faced to multi-hierarchy, user-oriented is presented. Such model can describe the geo-space on independent and complete entities, which are of geo-significance. Thus it is convenient for construction of complex geo-entities for different use, and it gears to the thinking custom of human beings, which makes it easy to be comprehended and accepted. So it can provide data- base for path planning, navigation and real-time updating of map data.It has been proved with experiments that methods in the paper accomplish the expected tasks well, which consolidates the base of recognition of map primitives as well as the construction and retrieval of data base.