摘要
形态特征是作物的肢体语言,对形态特征的研究在育种工作、生长状态分析、病害判断等方面都有着重要的意义。文章通过对图像分割和骨架提取算法进行优化改进提取出小麦植株骨架,采用最小矩形法计算小麦株高、用像素查找法计算叶长、用Hough变换计算叶基角和开张角,检测精度均达到90%以上,从而实现了面向农学小麦育种和管理专家的高效检测系统。
Morphologic characteristic is the body language of crops and it is of great value in the fields of breeding work, judgments of growth condition and disorder. Extracted skeleton by improving the image segmentation and skeleton extraction methods. Height was calculated by smallest rectangle method, leaf length was by pixel find method, leaf base angle were and out angle by Hough transform method. The inspected accuracy was all higher than 90%, achieved high efficiency detection system for agricultural wheat experts.
出处
《东北农业大学学报》
CAS
CSCD
北大核心
2009年第4期111-115,共5页
Journal of Northeast Agricultural University
关键词
小麦
形态特征
图像处理技术
wheat
morphologic characteristics
image procession
