A novel method to estimate fractal dimension of color images

Abstract

Fractal theory is one of the vital topics in image processing. The term Fractal Dimension (FD) is an important aspect in Fractal theory. The dimension of complex objects in nature is calculated by Fractal Dimension. FD is used in shape classification, graphic analysis and texture segmentation. FD also caters to several data mining tasks. One of the preferable method for calculation of FD of an image is box counting method. In this paper various methods used to calculate the FD of color images are emphasized. One of the existing method is discussed with its key problems and three methods has been proposed to achieve accurate FD. Experiments are conducted using these methods and FD is compared.