多幅图像显示时色彩压缩的聚类统计算法

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．

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

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.