基于openCV的玉米株型骨架提取研究  

Research on framework extraction software in maize plant type based on openCV

作　　者：刘同海[1] 白鹏[1] 李乃祥[1] 何玉香[1] 李少昆[2] 

机构地区：[1]天津农学院计算机科学与信息工程系,天津300384 [2]中国农科院作物科学研究所,北京100081

出　　处：《广东农业科学》2011年第20期163-165,共3页Guangdong Agricultural Sciences

基　　金：国家星火计划项目(2010GA610018);天津市农业科技成果转化与推广项目(0804110)

摘　　要：作物株型形态参数,一直是作物栽培及育种中所关注的指标。以粮食作物玉米的株形图像为研究对象,基于Intel公司的开源计算机视觉库OpenCV,以Microsoft Visual Studio 2008为开发平台,综合运用图像处理与分析技术,提出了一种适应玉米株形骨架提取的简便算法,并开发了玉米株型骨架提取软件。实证检验表明,该软件能实现载入一张复杂背景的玉米株形图像,经过剪切、自适应窗口等前期处理及对原始彩色图像的灰度化处理、图像分割和噪声去除等一系列的预处理,最后采用基于C#版的OpenCV(EmguCV)的一个算法提取出玉米株形的简单骨架。Morphological parameters of crop plant type are indexes that always been concerned by researchers in field of crop cultivation and breeding.In this paper, image of maize plant type was chosen as research object,Microsoft Visual Studio 2008 was sleeted as development tools,with the open source computer vision library （openCV）,we combined the image processing and analysis technology and propose a simple but powerful algorithm which was suitable to extract maize plant type framework.Experimental results showed that software developed with this algorithm could load a maize type image with complicated background,after a series of preprocesses such as cutting,self-adaptive window,graying of original colorful image,image segmentation,noise removal,a simple framework of maize type is extracted with our software developed with OpenCV（EmguCV） based on C#.

关 键 词：骨架提取 玉米株型 图像处理 OPENCV 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术] S513[自动化与计算机技术—计算机科学与技术]