Parallel object classification algorithms in images

Abstract

The contribution concerns  a parallelization of object classification algorithms for a SIMD-type parallel machine. It is assumed that gray-level values of image pixels are located in the orthogonal memory block on which a vector of one-bit processor operates in the bit-serial and   word – parallel mode. For this computer, the Bayes classification algorithm, the K-mean algorithm and  the ensemble average classifier and described.