METHOD FOR MAPPING GRAPHS ASSOCIATED WITH PHYSICAL ENTITIES AND ITS APPLICATIONS TO IMAGE REGISTRATION AND PATTERN RECOGNITION

摘要

methods 24 matching graphs associated with physical entities and its applications in image alignment and pattern recognition.the process of matching two graphs associated with physical entities, the two graphs containing a number of nodes n1 and n2 as n1. n2, isin the optimization process, to change a function of association observed repeatedly between nodes in order to minimize the two graphs by successive approaches.a function c (f) is called the matching cost function, defined as a measure of the error made by the association of current association function.application in the field of image analysis, and particularly in the matched images from different sensors and pattern recognition in three dimensions. figure 8.