Superpixel Image Segmentation Based on Fuzzy Clustering
Aiming at the problems of insufficient information retention and poor anti-noise properties of traditional fuzzy clustering algorithms in image segmentation,this paper proposes a super-pixel image segmentation based on fuzzy cluster-ing.The segmentation method first uses the Suppressed Fuzzy C-Means clustering algorithm to determine the classification of pixels in the global range.The boundary information of the object in the image is then fused at the local scale by using the Simple Linear Iterative Clustering.Finally,the image segmentation is completed by calculating the mean degree of affiliation under the superpixel segmentation grid division.This method not only retains the advantages of fuzzy clustering in coarse module segmentation,but also uses superpixel algorithm to improve the fusion ability of boundary information,so as to achieve the purpose of optimizing the segmentation effect.In the numerical experiment section,the effect of the algorithm is tested on synthetic images and the Berkeley data set respectively,and the property of boundary segmentation and anti-image noise is verified.
万方数据
维普
