A Sub-pixel Multifractal Method for the Image Segmentation

Abstract

The framework of image segmentation based on the sub-pixel multifractal measure (SPMM) is presented in this paper. A more precise singularity exponent distribution in the image can be obtained based on the SPMM. According to the singularity exponents and their statistical properties, the image can be decomposed into a series of sets with different physical characteristics automatically and easily. Moreover, the most singular manifold can be interpreted as the set from which energy is injected in the flow to the other fractal sets. The simulation results show that the SPMM has higher quality factor in the image edge detection.