Automatic recognition of topographic map symbols based on their textures

  • Authors:
  • Rudolf Szendrei;István Elek;István Fekete

  • Affiliations:
  • ELTE University, Faculty of Informatics, Budapest;ELTE University, Faculty of Informatics, Budapest;ELTE University, Faculty of Informatics, Budapest

  • Venue:
  • ICSI'11 Proceedings of the Second international conference on Advances in swarm intelligence - Volume Part II
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

The authors' research goal is to automatize the raster-vector conversion of topographic maps. To accomplish this goal, a software system is currently, under development. It separates the recognition of pointlike, line-like and surface-like objects. The first of these three topics is discussed in details in this paper. It is assumed that a topographic map and its vectorized form (possibly in rough form) are given. In this paper a method is introduced that is able to recognize the pointlike symbols of themap and to assign them as attributes to the corresponding polygon of the vectorizedmap. Itmeans that point-like symbols should not appear as polygons in the vectorized data model. Instead, symbols appear as polygon attributes. The method presented here is also able to "clean" unnecessary polygons of symbols from vectorized map. This method is implemented by optimized pattern matching on the raster image source of the map, where the symbols are handled as special textures. This method will be improved by using a raw vector model and a set of kernel symbols.