【摘要】 本文针对人脑CT图像的出血病灶区域,提出了一种改进的模糊C-均值(Fuzzy C-Means,FCM)算法进行颅脑内出血病灶的分割。首先对颅脑CT图像进行预分割,通过左右扫描算法和中值滤波算法将颅内结构从源CT图像中提取出来;然后对预分割而得到的颅内结构,利用在目标函数和隶属度函数中分别添加空间信息的改进FCM聚类算法进行出血病灶提取。通过对CT颅脑图像和添加椒盐噪声的CT颅脑图像进行病灶分割,结果显示本文算法对噪声不敏感,可以准确分割出出血病灶。更多还原

【Abstract】 In this paper, an improved fuzzy C-means(FCM) algorithm for the segmentation of intracranial hemorrhage lesions was proposed for the hemorrhagic lesions of human brain CT images. Firstly, the brain CT images were pre-divided, and the intracranial structures were extracted from the source CT images by left and right scanning algorithm and median fi ltering algorithm. Then the pre-segmentation intracranial structures were obtained by adding the objective function and membership function to the spatial information of improved FCM clustering algorithm for extraction of hemorrhagic lesions. Through CT brain images and CT brain images with salt and pepper noise segmentation, the results showed that the algorithm was insensitive to noise and can accurately segregate hemorrhagic lesions.更多还原