Geometrical Fusion Based on Chain Code Representation

Abstract

In this paper, a new geometrical image fusion method introduced by using the representation of chain code for the image objects that been matched among the images to be fused. The input images are separated into number of objects. A matching method is applied to extract the matching objects from the overall objects that included in the origin images. Then the chain code string is calculated for each one. The run length encoding is applied on that code strings to reduce the space that required. The image resulted from the geometrical fusion is more informative which contain the more important features those included in the origin images.