多幅图像显示时色彩压缩的聚类统计算法

A Clustering Statistics Algorithm For The Color Compression Of Images In Displaying Of Several Images

在计算机上，同屏显示多幅彩色图像会出现明显的失真（即色彩损失）．特别是两幅图像色调差别较大时（如一幅暖色调，一幅冷色调），失真现象更为严重．为了克服这一缺陷．本文在参考Ｋ均值聚类算法的基础上，优化了初始聚类中心，提出了色彩压缩的聚类统计算法．本算法简单实现较为方便，克服了Ｋ均值算法对色彩压缩不均衡及彩色失真度大的缺点．且不需经过多次迭代即可得到令人满意的效果，压缩时间短．

When displayes several images in same screen, it can be clear lack fidelity(color lose). When there is a marked defference of two images (for example: one is warm tone, the other is cold tone),the appearance of the lack fidelity is more seriously. In order to overcome thes shortcoming, thispaper optimizates initial clustering center and adcances a new algorithm of color compression clustering statistics algorithm on the basis of K-mean algorithm.This algorithm is simple and convenient. It overcomes the shortcoming of the K-mean algorithm, such as the unbalanced color compression, the lack fidelity of the color of the compressed images etc..It is a quite good algorithm of color compression now.