BINARY PARTITION ALGORITHMS AND VLSI ARCHITECTURES FOR MEDIAN AND RANK ORDER FILTERING

Abstract

A class of selection algorithms by binary partition is very efficient for median and rank order filtering. A unified discussion of these algorithms is presented. Binary partition algorithms have better time-area complexity than sorting methods. Counting, firing, and updating are three basic steps. A generic structure is proposed to realize these algorithms. They can be implemented by simple and regular modules in VLSI.