Region growing on a highly parallel mesh-connected SIMD computer

A parallel method for region growing on a highly parallel single-instruction multiple-data (SIMD) mesh computer is presented. The approach is based on a parallel merging paradigm, which involves the selection of the best of all merge possibilities for all regions concurrently. A key requirement of any parallel region growing scheme is the ability to compute functions concurrently on irregularly shaped regions. A set of general primitive functions for region growing are defined, and techniques to implement these functions on an SIMD processor are developed. These techniques make use of an embedded tree data structure to represent regions. The results of implementing a parallel split and merge region growing algorithm on the massively parallel processor are discussed. The approach is shown to be efficient primarily for images involving large numbers of regions