改进近邻传播聚类的彩色图像分割

针对近邻传播(AP)聚类算法存在运算时间长、空间复杂度高而难以应用于较大规模图像数据处理的问题,提出一种将mean shift(MS)算法和AP算法相结合的彩色图像分割方法——MSAP算法.首先应用MS算法对输入目标图像进行预分割,将分割后的区域数目代替原图像像素点数目作为AP算法输入数据的规模,计算每个区域中所有像素的彩色向量平均值,并将其作为AP算法输入的数据点,选用数据点间的距离作为相似度的测度指标；然后应用AP算法在数据相似度矩阵上进行聚类,得到最终的图像分割结果.实验结果表明,与AP算法相比,MSAP算法在运行时间和分割效果方面都有显著的提高.

Color Image Segmentation Based on Improved Affinity Propagation Clustering

The affinity propagation clustering algorithm requires huge storage and has high computational complexity.It is hard to be applied in image data real-time processing.A new color image segmentation method combining mean shift(MS) and affinity propagation(AP) named MSAP is presented in this paper.The proposed method preprocesses an input image with the MS algorithm.The numbers of segmented regions,instead of the numbers of image pixels,are considered as the input data scale of the AP algorithm.The average of the color vectors in each region is calculated as an input data point of AP algorithm.Distances between data points are regards as similarity measure indices,and then the AP algorithm is applied to perform globally optimized clustering and segmentation based on the similarity matrix.Experimental results illustrate that the MSAP method has superior performance and lower computational cost comparing with the AP algorithm.