Map Of The Road, The Identification Of River Information Extraction Method

It is one of the important research topics that automatic recognition of geographic maps combines the artificial intelligence and the digital image processing techniques to solve collecting, setting up the database and comprehending automatically the geographic space information in the Geographic Information System. And as basal geographic data, roads and rivers have always been the key link in such projects. In the last few years, some developments have been made about their extraction from map, but in regard to study condition, these aspects of their recognition rate, accuracy of recognition, automate level and maneuverability are problems of needing to resolve.The method of recognition and extraction the roads and rivers in colorful map was studied. According to the characteristic of the road and river, there were two methods to be proposed, the road extracted method: make use of the compactness and the color characteristics combine together, and a binary based on gray image extracting the rivers method. And there were some problem to be introduced, for example: the gray processing of image, the theories of mathematical morphology, binary, thinning and vectorization.Through the study, there are results that making use of this road extracted method is able to get good processed result and erase the characters noise of the road in the process of extraction; and this river extracted method is not only simple and fast, but also is able to realize the process of binary when the rivers are extracting.Finally based on the theoretic research accomplishment of this paper combined related program, the author developed an experimental system of the map information recognition in Visual C++ development environment. The experimental result of this system demonstrates that the two extracted methods are not only viable, but also obtain the extracted result well.