基于FCM聚类的自适应彩色图像分割算法

Adaptive color image segmentation algorithm based on FCM clustering

针对现有模糊c-均值聚类(FCM)算法存在的抗噪性能不佳且依赖初始条件的问题,提出一种基于FCM聚类的自适应彩色图像分割算法。通过直方图阈值方法分别得到彩色图像RGB各分量直方图的分割阈值,利用区域分裂合并方法获得聚类个数和初始聚类中心,使用一种考虑了像素间邻域信息的模糊c-均值聚类算法对图像聚类,得到最终分割结果。实验结果表明,该算法对彩色图像具有良好的分割效果,与现有同类主要算法相比,其分割效果和抗噪性能都有明显提高。

Aiming at the problems that the existing fuzzy c-means clustering（FCM）algorithms have poor anti-noise performance and they depend on the initial conditions,an adaptive color image segmentation algorithm based on FCM clustering was proposed.The thresholds of R,G and B histogram of color images were obtained using histogram thresholding method respectively.The number of clusters and the initial cluster centers were obtained using the region splitting and merging method.The fuzzy c-mean clustering algorithm based on spatial neighborhood pixels was used to cluster the image,and the final segmentation result was obtained.Experimental results demonstrate that the proposed algorithm has good segmentation effects on color images.Compared with the existing similar algorithms,the segmentation effect and noise resistance performance are improved obviously.