一种改进的快速FCM图像分割算法

FCM算法对图像的模糊特征具有较强的鲁棒性,在图像分割方面得到了广泛应用。但FCM算法采用随机初始化聚类中心的方法,使算法在迭代次数上有一定的不确定性。为提高FCM算法的运算效率,提出一种基于确定初始聚类中心的快速FCM图像分割算法。用最大类间方差法多次划分图像的灰度区间,根据区间中像素点的灰度值来初始化聚类中心,以使其尽可能的接近最终分割的聚类中心,减少算法的迭代次数。实验结果表明,与传统的FCM算法相比较,改进后的算法可以通过较少的迭代次数及运算时间分割图像。且该算法可以应用于诸多采取随机初始化聚类中心的FCM相关的算法中,以提高算法的运算效率。

An Improved Method of Fast Fuzzy C-means Image Segmentation

The FCM algorithm has strong robustness to the fuzzy features of the image and has been widely used in image segmentation. However,the FCM algorithm adopts a method of randomly initializing a cluster center,so that the algorithm has some uncertainty in the number of iterations. In order to improve the computational efficiency of FCM algorithm,in this paper,a fast FCM image segmentation algorithm based on determining the initial cluster cen-ters is proposed. The gray-scale interval of the image is divided by the maximum inter-class variance method,and the clustering center is initialized according to the gray value of the pixel in the interval. So that it is as close as possible to the final clustering of the clustering centers,thereby reducing the number of iterations of the algorithm. The experi-mental results show that compared with the traditional FCM algorithm,the image can be segmented by the improved algorithm with fewer iterations and the computation time. The algorithm can also be applied to many FCM-related al-gorithms which take random initialization clustering center to further improve the operating efficiency of the algorithm.