Abstract:
Based on the analysis of common segmentation methods,a fuzzy C-means (FCM) clustering algorithm with superpixel is proposed in this paper.Firstly,a kind of pixel similarity considering intensity and distance is defined and the image superpixels are generated.After that,the wavelet energy features are extracted from the superpixels,which are clustered by FCM.The experimental results demonstrate that the proposed method is robust to noise and isolated pixels,which can improve the segmentation accuracy.