基于Mean-Shift聚类的色彩传递算法  

Color Transfer Algorithm Based on Mean-Shift Clustering

作　　者：张二虎[1] 张洋[1] 

机构地区：[1]西安理工大学印刷包装工程学院,陕西西安710048

出　　处：《西安理工大学学报》2009年第1期105-109,共5页Journal of Xi'an University of Technology

基　　金：陕西省科技发展计划项目(2008K07-20)

摘　　要：针对当前色彩传递算法存在的传递效果不准确的问题,提出了一种基于Mean-Shift聚类的色彩传递算法。首先将输入图像和样本图像从RGB空间转换到LUV空间,在LUV空间对2幅图像进行Mean-Shift模糊颜色聚类分析;根据输入图像聚类域的统计信息,在样本图像中找到一个与统计信息最为接近的域作为该聚类域的匹配域;最后将隶属度向量引入色彩传递方程,实现图像间自动的色彩传递。实验结果表明,该算法较好地实现了彩色图像之间的色彩传递和灰度图像的上色处理,无需设定手动聚类数目,并且迭代过程较简单,色彩传递效果更好。With an aim at the inaccurate problem of existing transfer effect by the current color transfer algorithm, this paper proposes a new color transfer method based on mean-shift fuzzy color clustering. Firstly, the input image and example image should be transformed from RGB color space to LUV color space; mean-shift fuzzy color clustering method is used to analyze two images in LUV space. Secondly, according to the local statistical properties, the cluster area of input image is matched with the most similar cluster area of example image. Finally, membership degree control is introduced to color transfer equations to realize the adaptive color transfer between images. Experimental results demonstrate that the proposed algorithm can better achieve the color transfer among the color images and painting treatment of gray-scale images without setting initial number of cluster and with simple iterative process and better color transfer effects.

关 键 词：色彩传递 均值平移 模糊聚类 图像处理 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]