We describe a widely applicable method of grouping or clustering image features (such as points, lines, corners, flow vectors and the like). It takes as input a "proximity matrix" H a square, symmetric matrix of dimension N (where N is the number of features). The element i,j of H is an initial estimate of the "proximity" between the ith and yth features. As output it delivers another square sy...

A construction is given for associating to any symmetric (v, k, X) design a self-dual code of length v+ 1 over GF (p), where p is any divisor of the square-free part of k — k. These codes are then used to obtain a substantial strengthening of a result of Hughes on automorphisms of designs. Specifically, suppose a symmetric {v, k, X) design possesses an automorphism a of odd prime order q. If an...

The maximization through signal design of the symmetric capacity of CDMA systems with optimal and minimum mean-square error decision feedback (MMSE-DF) decoding is considered. In particular, lower bounds on maximum symmetric capacity (and the signal set that achieves them) are derived for the two decoders. A common sufficient condition is obtained for both decoders under which the signal design...

The power index of a square Boolean matrix A is the least integer d such that A is a linear combination of previous nonnegative powers of A. We determine the maximum power indices for the class of n × n primitive symmetric Boolean matrices of trace zero, the class of n × n irreducible nonprimitive symmetric Boolean matrices, and the class of n×n reducible symmetric Boolean matrices of trace zer...