A Parallel Solution in Texture Analysis Employing a Massively Parallel Processor

Abstract

Texture is a fundamental feature for image analysis, classification, and segmentation. Therefore, the reduction of the time needed for its description in a real application environment is an important objective. In this paper, a texture description algorithm running over a hypercube massively parallel processor, is presented and evaluated through its application in real texture analysis. It is also shown that its hardware requirements can be tolerated by modern VLSI technology.