计算机工程 ›› 2009, Vol. 35 ›› Issue (19): 203-205.doi: 10.3969/j.issn.1000-3428.2009.19.068

• 图形图像处理 • 上一篇    下一篇

基于小波变换和KFCM的彩色图像分割

李志梅1,2,肖德贵2,王丽丽1   

  1. (1. 桂林航天工业高等专科学校计算机系,桂林 541004;2. 湖南大学计算机与通信学院,长沙 410082)
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2009-10-05 发布日期:2009-10-05

Color Image Segmentation Based on Wavelet Transform and Kernelized Fuzzy C-Mean

LI Zhi-mei1,2, XIAO De-gui2, WANG Li-li1   

  1. (1. Department of Computer Science and Technology, Guilin College of Aerospace Technology, Guilin 541004; 2. School of Computer and Communication, Hunan University, Changsha 410082)
  • Received:1900-01-01 Revised:1900-01-01 Online:2009-10-05 Published:2009-10-05

摘要: 提出一种将小波变换和核模糊C均值聚类算法相结合的快速彩色图像分割算法。利用小波变换的多分辨率特性,在分辨率最大尺度上的LL子带进行均值漂移聚类,快速获得初始粗分割结果,在其基础上进行模糊核聚类分割,将上一层的结果用于下一层的初始化,重复至最低分辨率后用最小分类器对原始图像进行最终分割。实验结果证明,该算法分割速度快,对自然彩色图像的分割结果优于模糊C均值算法和均值漂移算法。

关键词: 小波变换, 图像分割, 核模糊C均值聚类, 均值漂移

Abstract: This paper proposes a fast colore image segmentation algorithm based on wavelet transform and Kernelized Fuzzy C-Mean(KFCM) clustering algorithm. By using the multiresolution of wavelet transform, mean-shift clustering is implemented on the LL sub-band of maximum size and the initial coarse image segmentation is obtained. Based on the initial coarse segmentation, KFCM is implemented and the result of the previous level is used to initialize the next level until the lowest resolution appears. The initial image is segmented with the minimum classifier. Experimental results demonstrate that this method has fast speed and performs better than traditional Fuzzy C-Mean(KFCM) algorithm and mean-shift algorithm.

Key words: wavelet transform, image segmentation, Kernelized Fuzzy C-Mean(KFCM) clustering, mean shift

中图分类号: