Abstract: A clustering and assignment processor (40) for correlating set (12) of two-dimensional data points (14). Clustering is performed by treating each data point (14) as a mass subject to gravitational attraction by other data points (14). The data points (14) are permitted to move in data space until all data points (14) in particular regions (16) of the data set (12) fall into a single point, or centroid (18). Individual data points (14) and correlated regions (16) of successive data sets (12) are then correlated by means of a resonant descent technique, which determines the distance and direction of individual data points (14) from a chosen data point (108). The processor (40) then matches a pair of two data points (12) with the chosen data point (108) which are equidistant and collinear with the chosen data point (108).