摘要
将微粒群算法应用于彩色图像的颜色量化.根据种群的进化原理,在RGB彩色空间中对图像像素的颜色值进行聚类,通过迭代寻优确定最优调色板,然后进行像素的颜色映射,从而完成量化工作.仿真实验表明,该算法具有颜色失真小,收敛性好,鲁棒性强和受初始条件影响小等特点,是一种有效的颜色量化算法.
The particle swarm optimization is applied color quantization of color image in this paper. According to the evolution theory of population, color values of the image pixels are clustered in the color space of RGB. Thresh iterative optimization, the most superior palette is determined, and quantization of color mapping of every pixel is completed. Experiment results show that the algorithm has the advantages of less distortion, good astringency, and excellent robustness.
出处
《微电子学与计算机》
CSCD
北大核心
2008年第3期51-54,共4页
Microelectronics & Computer
基金
湖南省自然科学基金项目(06JJ50110)
湖南省教育厅科学研究项目(04C315)
关键词
微粒群
颜色量化
聚类
particle swarm optimization
color quantization
clustering
