A complex wavelet based image segmentation using MKFCM clustering and Adaptive level set method

In this paper, a novel image segmentation algorithm is proposed which combines the Dual tree complex wavelet transform (DT-CWT), Multiple kernel fuzzy c-means clustering (MKFCM) and Adaptive level set method. The Dual tree complex wavelet transform is used for image denoising. Also it extracts high frequency components of image where in wavelets representation of image details is presented in high frequency subbands. After denoising the noisy image multiple kernel fuzzy c-means clustering algorithm is applied to separate an image into number of homogeneous non overlapped closed regions. Also this algorithm computes the fuzzy membership values of each pixel. Based on Multiple kernel fuzzy c-means clustering edge indicator function was redefined. Then Adaptive level set method is applied to extracting the boundaries of objects on the basis of the MKFCM segmentation. The efficiency and accuracy of the proposed algorithm is shown by experimenting on the noisy MRI and white blood cell images.