计算机视觉技术在作物形态测量中的应用  

Application of technology for computer vision in plants shape measurement

作　　者：徐歆恺[1] 郭楠[1] 葛庆平[1] 郭新宇[2] 

机构地区：[1]首都师范大学信息工程学院,北京100037 [2]国家农业信息化工程技术研究中心,北京100089

出　　处：《计算机工程与设计》2006年第7期1134-1136,1148,共4页Computer Engineering and Design

基　　金：国家863高技术研究发展计划基金项目(2001AA115190-08);北京市教委基金项目(KM200310028109)

摘　　要：为了测量作物形态信息,制定了玉米植株照片拍摄标准;利用对玉米图像细化后叶片和茎杆间存在的交点数目来检测玉米叶片的数量,采用噪声和毛刺去除算法提高了系统检测的精度;改进了freeman链码的长度测量方法,大大提高了曲线长度的测量精度。用Microsoft Visual C++.NET程序开发工具实现了作物形态无损检测系统,通过实验表明,该系统可以对作物叶长、叶倾角等形态特征进行自动和手动的测量,精度符合设计要求。The system is developed to measure the shape of plants. Several improvements have been made in this software system： One is that the standard for taking photo of maize is worked out, in the thinned maize photo, the number of the maize leaves is calculated by counting the joining-point between the leaf and the stalk, the new noise-killing and blur-thinning algorithms make measurement more accurate, improved freeman code length measuring method improve the precision of the curve-length, harmless plant-shape measuring system, created by microsoft visual C＋＋.net, can measure the plant＇s shape in aspects like the leafs length and stem-leaf angle either manually or automatically, and the precision accords with the designing requirement.

关 键 词：计算机视觉 作物形态 测量系统 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]