Reduction of resolution for binary images by an octal grid graph representation model

We deal with a data structure which is suitable for expression of the binary images. In general, binary images are expressed in raster data or rectangular dissections. In this paper, we describe that it is possible to reduce effectively the number of nodes in raster data represented by “octgrid” which is a type of octal grid graphs to handle heterogeneous rectangular dissections. Precisely, we propose algorithms to reduce the resolution of binary images represented by “octgrid”. Furthermore, we compare by examples our method with the Hilbert scanning, and show that our method reduces the rectangular number by less than 50% than the Hilbert scanning does, in several examples.