Algorithm for Image Integration Invariant to Disproportional Scaling

Imagery integration has a variety of open questions in image registration, conflation, and search. A new promising approach is based on the algebraic paradigm that differs from the more traditional geometric and topological approaches. In previous research, algebraic methods have been based on algebraic structural analysis of features represented as open polylines called linear features. In this paper we report a new algebraic method based on ratios of areas of closed contours (shapes). It has been proven that this method is invariant to disproportional scaling. The method is applicable to images that contain three or more matching shapes, but those shapes and their actual matches are not known. The method has also undergone some optimization to shorten run time. Computational experiments had shown that the method is capable of integrating images with dozens of shapes in a reasonable time. The algorithm is also a result of the algorithm development technology proposed in this study. The method is implemented as an ArcMap Plug-in.