基于HSI色彩空间与FFCM聚类的葡萄图像分割  

Segmentation of Grapes Image Based on HSI Color Space and FFCM Clustering Algorithm

作　　者：宋西平[1] 李国琴[1] 罗陆锋[1,2] 邹湘军[2] 张丛[1] 

机构地区：[1]天津职业技术师范大学机械工程学院,天津300222 [2]华南农业大学工程学院,广州510642

出　　处：《农机化研究》2015年第10期40-44,共5页Journal of Agricultural Mechanization Research

基　　金：国家自然科学基金项目(31171457)

摘　　要：为适应农业采摘机器人对葡萄对象快速准确识别的需要,提出了基于HSI色彩空间与以直方图信息为特征的快速模糊C-均值聚类(FFCM)算法相结合的葡萄图像分割方法。该方法以H分量作为葡萄图像聚类分割的处理数据,根据FFCM算法对灰度图像聚类分割。试验对夏黑葡萄果实在自然光、顺光、背光照射环境下拍摄的图像进行分割。结果表明:葡萄图像分割方法能够快速且较好地从复杂自然环境中将葡萄目标分割出来,为葡萄采摘机器人的研制提供了重要参考。In order to adapt to the agricultural picking robots needing to recognize grapes quickly and accurately , a new segmentation method that based on HSI color space and fast fuzzy C-means clustering （ FFCM ） algorithm characterized by histogram informations is presented .This method is that original RGB images is transformed into HSI images , and using H component as clustering segmentation of data to cluster gray image according to the FFCM algorithm .By analying the testing of xiahei grapes under natural light , facing of the light and backlight ,the image segmentation method of this paper can split grapes fastly from the complex natural environments and provide some references for the research of agri -cultural robots .

关 键 词：葡萄 图像分割 HSI色彩空间 直方图 快速模糊C-均值聚类 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] S663.1[自动化与计算机技术—计算机科学与技术]