Texture segmentation using multifractal measures

This paper presents a study of application of multifractal measures of grey-level images through the generalized Renyi entropy. Grey-level images are analyzed from the point of view of strange attractors. This paper shows that the singularity dimension in the multifractal measures can effectively reflect the nonuniform property of the image. Different textures can be separated because similar textures generally have homogeneous properties which can be characterized by the singularity and Mandelbrot spectra of the fractal sets. By taking the rate of change of the singularity, better image segmentation has been achieved. The advantage of this technique over alternative classical operators can be seen fully when it is applied to some very complicated images such as malignant cancer cell images