Object Characterization in Greyscale Imagery using Fractal Dimension

摘要

Local fractal dimension operators can be used to discriminate among objects ingreyscale imagery. The mass scaling (box-counting) algorithm for computing local fractal dimension promises real-time results when screening numerous or large images. In this paper we show the ambiguities in the mass scaling technique based on shape and range, we examine previously published algorithms for computing local fractal dimension, we show by example some of their deficiencies, and we propose a new algorithm that solves some of these problems.