A Novel Image Segmentation Algorithm for Clinical CT Images Using Wavelet Transform, Curvelet Transform and Multiple Kernel FCM

The clarity of medical image, which is directly acquired from the scanning machine, is very less. Image enhancement is one of the best and efficient techniques to increase the quality of image. A combined approach of different techniques such as Wavelet, Curvelet and Multiple Kernel Fuzzy C-Means algorithm was carry out in this paper. Wavelet and Curvelet transforms are used for denoising purpose. Due to wavelet transform’s excellent localization property, it is more suitable for denoising the homogeneous areas of the image. Curvelet transform is a new multiscale representation and it is most suitable for the objects with curves. It is a new extension of the wavelet transform and ridge let transform and preferred for two dimensional images. Multiple Kernel Fuzzy C-means (MKFCM) algorithm is used for segmentation purpose of the image. Parameters such as mean, standard deviation, entropy and peak signal-to-noise ratio are used to measure thesegmentation efficiency. From the experimental results it is clear that the proposed segmentation technique produces maximum efficiency and is suitable for the segmentation of Clinical CT images. The main advantages of the proposed technique are simplicity, reliability and fast convergence.