基于视觉特性的一种自适应颜色量化算法

摘要

Color quantization is an important technique of digital image process. There are two main algorithms for color quantization, the cutting algorithm and the clustering algorithm. This paper presented an self-adaptive color quantization algorithm based on visual characteristic of human being. The experiment results showed that the algorithm we proposed has a good visual effect of images which is very close to that of K-mean and a high time processing performance which is very close to that of Mean-cut.%　　颜色量化是数字图像处理的基本技术之一，传统的量化算法有分割法和聚类法。本文提出了一种结合分割法和聚类法特点的自适应颜色量化算法，实验结果表明，在量化质量上本算法接近 K-均值聚类法，但在运行效率上十分接近中位切分法，是一种相对高质高效的颜色量化算法。