MAP STRUCTURE RECOGNITION AND AUTOMATIC GENERATION OF DIGITAL ROAD MAP AND HOUSE MAP DATABASES

In this paper, an experimental system of automatic digital map database generation is presented. Taking pictorial maps as input data, the system can generate map data of roads, houses and map contour lines. The system consists of several software modules: preprocessing, map image thinning, map graph conversion, map graph making-up, map component separation, road map recognition and vectorization, house map recognition and polygonalization, map data management and interactive data correction. The operations of preprocessing and thinning are performed in pixel image domains, while map graph conversion converts skeletons of map images to map graphs. All the processes afterwards are then carried out in graph domains. Map graph making-up repairs imperfections of maps, such as connecting the gaps of line draws that might inherently exist in the original master maps. Map component separation decomposes the connectivity of different map components. The processes of road map recognition and vectorization detect road structures from map graphs and generate vector data of roads, while house map recognition and polygonalization identify house structures and generate vector data of houses. The module of map data management is responsible for map image partition, map data rejoining and map data formatting. Interactive data correction is a computer interface that is used for data error corrections. The road data and house data produced by the system are in a listed link format, which is usually used in vehicle navigation, vehicle location and land planning. For the covering abstract see IRRD E102946.

Language

  • English

Subject/Index Terms

Filing Info

  • Accession Number: 00781105
  • Record Type: Publication
  • Source Agency: Transport Research Laboratory
  • ISBN: 89-950073-2-X
  • Files: ITRD
  • Created Date: Jan 7 2000 12:00AM