K-means聚类数的确定及在细胞图像颜色校正中的应用  

Determination of Number of Clusters in K-means and Application in Color Correction of Cell Image

作　　者：罗丽丽[1] 蔡坚勇[1,2,3,4] 蔡荣太[1,3] 林李金[1] 蔡娟[1] 

机构地区：[1]福建师范大学光电与信息工程学院,福州350007 [2]福建师范大学医学光电科学与技术教育部重点实验室,福州350007 [3]福建师范大学福建省光子技术重点实验室,福州350007 [4]福建师范大学智能光电系统工程研究中心,福州350007

出　　处：《计算机系统应用》2014年第1期114-118,共5页Computer Systems & Applications

基　　金：福建省高校产学合作科技重大项目(2011H6010);国家自然科学基金(61179011)

摘　　要：针对大量瑞氏染色细胞图像,通过YCbCr颜色空间进行K-means聚类,观察各分量聚类中心差值变化规律,提出了一种新的确定K-means聚类数的颜色校正算法.该算法首先是将瑞氏染色细胞图像中不同目标分别准确地聚集在相应类当中,再与标准图像中的每类进行配比,并利用直方图规定化进行直方图调整,得到颜色校正结果.经大量实验证明,尤其在细胞图像中目标颜色特征较接近的情况下,该算法通过确定合适的聚类数可大大提高颜色校正结果的准确率.To observe the changing rule of clustering center’s value in K-means by the YCbCr color space, against a lot of cell image by Wright Stain, this paper proposed a new method to determine the number of clusters in K-means to get a good result of color correction. Different target of cell image could be firstly gathered in the corresponding class accurately, then by matching between classes, and using histogram specification for adjusting histogram, the results of color correction could at last be achieved. It has been proved by many experiments that this algorithm can greatly improve the accuracy of color correction results, especially the target with closer color features in cell.

关 键 词：K-MEANS 中心差值 聚类数 颜色校正 准确率 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]