【摘要】 针对现有模糊C均值(FCM)聚类算法存在的对初始参数敏感、迭代速度慢,以及对噪声鲁棒性差等问题,提出将蚁群优化算法(ACO)与直觉模糊聚类相结合的方法用于分割脑部MR图像。该算法采用自适应蚁群优化算法获取初始聚类中心与聚类个数作为直觉模糊聚类的初始值,将融入了局部空间信息和犹豫度的直觉模糊聚类算法应用于含噪声脑部图像及脑肿瘤图像进行分割。实验结果表明该算法能够有效抑制噪声干扰且保存图像细节,相较于FCM及相关改进算法具有更高的分割精度和分割效率。更多还原

【Abstract】 Aimed at the problems of the existing fuzzy C-means(FCM) clustering algorithm such as being sensitive to initial parameters, slow iteration speed, poor robustness to noise and so on, the method for combining the ant colony optimization(ACO) algorithm with the intuitionistic fuzzy clustering is proposed for segmenting the MR images of the brain. The proposed adaptive ant colony optimization algorithm was used to obtain the initial value of the initial clustering center and the number of clusters as the initial value of the intuitionistic fuzzy cluster, and then the intuitionistic fuzzy clustering algorithm, which was integrated into the local spatial information and the hesitation degree, was applied to the segmentation of the noise-containing brain image and brain tumor image. The experimental results show that the algorithm can effectively suppress the noise interference and save the image detail, and has higher segmentation precision and segmentation efficiency compared with the FCM and the related improvement algorithm.更多还原