Statistical Multiscale Image Segmentation via Alpha-Stable Modeling

This paper presents a new statistical image segmentation algorithm, in which the texture features are modeled by Symmetric Alpha-Stable (SαS) distributions. These features are ef ciently combined with the dominant color feature to perform automatic segmentation. First, the image is roughly segmented into textured and nontextured regions using the DualTree Complex Wavelet Transform (DT-CWT) with the subband coef cients modeled as SαS random variables. A multiscale segmentation is then applied to the resulting regions, according to the local texture characteristics. Finally, a novel statistical region merging algorithm is introduced by measuring the Kullback-Leibler distance (KLD) between estimated SαS models for the neighboring segments. Experiments show that our algorithm achieves superior segmentation results in comparison with existing state-of-the-art image segmentation algorithms.

• Type: Conference
• Year: 2007
• Where: ICIP
• Authors: Tao Wan, Cedric Nishan Canagarajah, Alin Achim
• Keywords: Algorithms, ICIP 2007, Image Processing, Image Segmentation, Statistical Image Segmentation