改进的模糊C-均值聚类医学图像分割算法

针对模糊C-均值聚类算法分割图像时容易产生模糊边缘的缺点,提出了一种结合图像梯度和模糊C-均值聚类的图像分割方法.该方法利用图像梯度反映出来的目标边界,对由模糊C-均值聚类所获得的聚类区域进行分割,把因模糊性而划分到目标区域的像素点与目标区域进行分离,同时利用区域增长方法找出干扰区域并删除.将该算法应用到胰腺ERCP图像分割,实验表明,改进算法能够比较准确地分割出图像中的目标,减少因模糊聚类产生的模糊边缘.

Improved fuzzy C-means clustering segmentation algorithm of medical image

The fuzzy C- means clustering image segmentation algorithm is easy to generate fuzzy edges. To overcome this im- age segmentation problem, a new method was proposed. It combined image gradient and the fuzzy C- means clustering image seg- mentation methods. It uses the target boundary reflected by the image gradient to segment the clustering region which was obtained by the fuzzy C- means clustering image segmentation methods, and then the target area and the target region were separated. At the same time using region growing method to find and remove interference region. Finally the algorithm is applied to the medical image segmentation, the experiments show that, the improved algorithm can accurately segment the image of the target arid reduce the fuzzy edges generated by fuzzy clustering.