This paper aims at a system which receives the image information of a commercially available map, separates the information according to their categories, and stores the information so that re‐editing is possible. Algorithms are proposed by which the map information given as the input image is separated with the least intervention of human operation, and only the incomplete points are indicated by the human observation. An evaluation of the algorithms is presented. In the past, various algorithms have been proposed for the image separation. In this work, an experimental system was constructed to demonstrate the feasibility of constructing the actual system. By the proposed algorithm, most of the characters, roads, coastlines, railways and boundary lines were separated from the black‐and‐white map. The information compression for storage and indexing for retrieval are also performed. The data storage method is discussed. Finally, the map output of the system is described, and several examples of output maps are presented.
ASJC Scopus subject areas