Image segmentation framework using wavelet transform

One of the most important tasks of image processing is the image segmentation where it should be extracted the significant features for the following classification or detection stages. Numerous methods have been developed in the segmentation, but by now, this problem demands finding efficient and fast algorithms. In this paper, we propose a novel framework that is based on cluster segmentation in the wavelet transform space, where wavelet decomposition procedure permits to analyze texture features in high frequency bands. Different simulation results, which have been revised in objective and subjective matter, confirm the efficiency of the novel segmentation framework for images of different nature, in particular dermoscopic images.