Character grouping technique using 3D neighborhood graphs in raster map

The main problem in this paper is how to find the character which is placed on a line or curves of raster map. We give one novel algorithm to group each separated characters in a map. This word grouping is difficult especially in a map, since a map has many different types of character and each word has its own slanting line. For this, we propose the 3D neighborhood graph G from a given set of characters. In this graph, each vertex of G represents the separated characters and it is placed in 3D space according to the size of the character. This makes the bigger characters being located in the upper position, the smaller characters being placed in the bottom. We give an edge if two vertices are nearly placed in that 3D space. By this edge connection strategy we can easily find the words of various different size in a map.