| With the rapid development of Medical Imaging Technology over these years, clinicians can observe the lesion position of patient more directly and clearly through these technologies, so as to make more accurate judgment and contribute to the treatment of patients. Using medical image segmentation technology can help us to analysis the shape, the border and the size of sick organs more clearly. Even computer can automatically analysis the segmentation result, and make preliminary judgment of the illness, so that people can reference. Due to individual differences in the human body, medical imaging classification algorithm in clinic requires a high segmentation precision and speed. At present, the segmentation method currently used by many, but is not perfect. Therefore, medical image segmentation algorithm is still a hot research.Image segmentation is a technology that it divides the image into several regions. The interior of each region has a high similarity among pixels, while different regions have a low one. The interference and influence of various factors in practice lead to the uncertainty of image segmentation, however, fuzzy theory and fuzzy image segmentation technology is suitable for this kind of uncertainty, so the focus of this paper is to apply the fuzzy clustering method in medical image segmentation. Typical method of fuzzy clustering is Fuzzy C-means (FCM), in fact it is a nonlinear programming problem with constraints, and fuzzy partition of the image is obtained by solving the optimization of the objective function. Based on the in-depth study of FCM algorithm theory, we will serve improving the algorithm of the selection of the initial clustering center, and the ability of noise processing as the research point. This paper introduces a new algorithm based on the kernel function that rationally uses the image space information and effectively solves the problem that the kernel Fuzzy C-means is not sensitive to the noise, and improves the segmentation accuracy. |
PDF Full Text Request