Analytical Study on Fractal Dimension- A Review

Fractal theory can be viewed as one of the vital aspects of image processing. The roughness and pattern details for any naturally occurring element can be described through Fractal Dimension (FD). FD proves to be an effective tool for classifying different shapes in texture segmentation and performing graphical analysis in various fields of image processing. The literature has set forth quite a few approaches for calculating FD. One of the most elegant method for gray images is box counting. This paper offers an analysis of the existing approaches applied on gray images for calculating their FD. The prime issues associated with these approaches are reviewed and finally a relative study of their complexity is showcased. We have also experimentally compared our previously proposed approach (Relative Improved DBC) with the improved DBC method. In addition, we have also established the difference among the FDs calculated by the existing methods in literature.