改进FCM算法在医学图像分割的方法研究

The Research of Modified Fuzzy C-Means Clustering Algorithm On Medical Image Segmentation

本文提出使用改进模糊C均值聚类(MFCM)算法和模糊可能性C均值聚类(FPCM)算法的图像分割方法并应用于医学图像分割过程中。MFCM算法是通过调整FCM算法的测量距离来批准标签像素受到其他图像像素和在切分中抑制噪声效果来约束,从而使得成员变量没有最大约束值。基于真实医学图像的实验表明了MFCM算法和FPCM算法在医学图像中进行分割的实际效果,具体是通过对FCM、MFCM、FPCM进行精度对比来验证算法有效性。

The paper proposes the Modified Fuzzy C-Means （MFCM） algorithm and fuzzy C-Means clustering （FPCM） algorithm for image segmentation and its application in medical image process. MFCM algorithm is adopted to adjust the traditional FCM algorithm to measure the distance to approval by other labels pixel image pixels and noise suppression effect in syncopation to constraint, so that the member variable no maximum constraint. The real image experiments show that the Modified FCM algorithm in medical image segmentation results, specifically through the traditional FCM, Modified FCM, FPCM accuracy compared to verify the algorithm validity.