A method of matching labeled point patterns is described on the basis of a fuzzy relaxation. The method is applicable to labeled point patterns which differ by geometrical transformations, such as translation, rotation and scale change. In this method, the point pairs are considered to be the pattern primitives. The pattern primitives are geometrically transformed in the process of relaxation so as to minimize a measure of mismatch between the primitive pairs. The compatible primitive pairs between the two labeled point patterns are established after only a few iterations of relaxation and the corresponding points are obtained from the primitive pairs. As an example, the method is applied to the matching of constellations. © 1984.
Ogawa, H. (1984). Labeled point pattern matching by fuzzy relaxation. Pattern Recognition, 17(5), 569–573. https://doi.org/10.1016/0031-3203(84)90055-4