基于模糊C均值的桉树茎切片图像纤维细胞区域提取技术

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

基金项目: 华南农业大学校长基金资助项目(5600k04099)

Research on extracting the fibre cell region in a eucalyptus image based on fuzzy C-means clustering

摘要: 介绍了一种自适应提取桉树茎切片图像中纤维细胞区域的图像处理技术．先制作桉树茎切片，通过数码显微镜获取桉树茎切片彩色图像．将彩色图像转换为CIE L^*a^*b^*彩色空间，用模糊C均值聚类算法(FCM)对a^*、b^*通道进行模糊聚类分析，计算出各种颜色细胞的聚类中心，以各个象素a^*、b^*分量的值与各个聚类中心之间的欧氏距离来区分各种细胞，然后提取出纤维细胞区域．结果表明，在CIE L^*a^*b^*空间使用FCM方法是提取桉树茎切片图像纤维细胞区域的有效方法．
- 彩色图像处理 /
- 纤维细胞 /
- 区域提取 /
- 模糊C均值聚类
Abstract: A method of extracting the fibre cell region from a eucalyptus image was introduced. After making the fibre cell slices, the color images were taken with a digital microscope. CIE L~*a~*b~* was utilized to conduct image processing for color images which satisfied extensive color ranging. The a~* and b~* are selected and clustered by fuzzy C-means (FCM). After the clusting center of various color cells was calculated, cells were distinguished from the euclid distance. An adaptive segmentation method using FCM clustering algorithm was employed to take out the fibre cell of eucalyptus.
- color image processing /
- fibre cell /
- region extract /
- fuzzy C-means cluster