基于图像处理的株高无损测量方法研究  

Measurement Method of Plant Height Based on Image Processing

作　　者：何晶[1] 

机构地区：[1]重庆青年职业技术学院信息工程系,重庆400712

出　　处：《测控技术》2015年第4期39-42,共4页Measurement & Control Technology

摘　　要：为了在定量化研究植物株高过程中实现无损测量,同时降低户外植物生长监测的硬件成本,方便设备安装和操作,依托立体视觉技术,提出了一种基于图像处理的株高无损测量方法。以测量棉花株高为研究对象,首先,在棉花生长无损监测系统中,通过标定后的单个摄像机获取棉花植株图像,在二维图像空间将棉花株茎简化成两个关键点坐标,然后利用摄像机参数矩阵以及对两个关键点的约束条件,获得植株在世界坐标系中的三维信息,从而获得植株高度。实验结果表明,该算法在进行植株高度测量时,其误差范围可以控制在5mm以内,该方法是可行的、有效的。In order to achieve non-destructive measurement in the process of quantitative research on plant height and also reduce hardware costs of outdoor plant growth monitoring,be easy to install and operate for equipment,relying on stereo vision technology,a non-destructive measurement method for plant height based on image processing is proposed.Measuring plant height as the study object,first,in the cotton growing non-destructive monitoring system,the cotton plant image is obtained by a calibrated single camera.In the two-dimensional image space,cotton stems are simplified into two key coordinate points.Then the three-dimensional plant information in the world coordinate system and the plant height are obtained by using the constraints of the camera parameter matrix and the two key points.Experimental results show that when the algorithm measuring plant height,its error range can be controlled in less than 5 mm,the method is feasible and effective.

关 键 词：株高测量 图像处理 立体视觉 无损检测 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]