基于K均值聚类的厚壁组织区域自动提取  

Research on Extracting the Sclerenchyma Region in a Eucalyptus Image Based on K-means Clustering

作　　者：王琴[1] 王卫星[2] 张少翃[3] 洪添胜[1] 俞新华[4] 

机构地区：[1]华南农业大学信息学院,广州510642 [2]华南农业大学工程学院,广州510642 [3]华南农业大学设备资产管理处,广州510642 [4]华南农业大学生命科学学院,广州510642

出　　处：《农机化研究》2006年第10期197-199,共3页Journal of Agricultural Mechanization Research

基　　金：华南农业大学校长基金(5600-k04099)

摘　　要：桉树是我国南方的速生丰产林,区域提取是定量分析各种组织抑制桉树茎生根原因的重要前提。为此,介绍了基于K均值聚类的在桉树茎切片图像中自适应提取厚壁组织区域的图像处理技术。将桉树茎切片彩色图像转换为CIEL*a*b*彩色空间,用K均值聚类分析算法对描述颜色的a*和b*通道进行聚类分析,提取细胞厚壁,然后填充其中白色的细胞腔,构成完整的厚壁组织区域。实验结果表明,在CIEL*a*b*空间使用该算法可以获得较准确的实验结果。A The Eucalyptus is a fast-growing and fertility tree in south China. The region extraction is very important to analyze quantification ally which kind of organization suppresses the Eucalyptus stem to take root. An image processing method of extracting the sclerenchyma Region from a Eucalyptus Image was introduced. The CIE L^＊a^＊b^＊ was utilized to conduct image processing for Eucalyptus slice Color images which satisfies extensive color ranging. The a^＊ and b^＊ are selected and clustered by K-means. After extracting the thick cell wall, the white cell cavity was filled in and the Sclerenchyma Region was formed. An adaptive segmentation method using a K-means clustering algorithm was employed to take out the sclerenchyma Region of eucalyptus.

关 键 词：林业基础科学 厚壁组织区域自动提取 应用 数字图像处理 Kmeans均值聚类 

分 类 号：S712[农业科学—林学] TP242.62[自动化与计算机技术—检测技术与自动化装置]