摘要
为了实现玉米幼苗叶面积的快速、无损、实时、高效检测,设计并搭建了基于机器视觉的玉米幼苗叶面积检测装置。该检测装置由框架、光源装置、顶升旋转系统、图像采集及分析系统、检测装置控制系统等部分组成,通过各部分协作完成玉米幼苗顶视图像与侧视图像的实时采集及分析处理,计算玉米幼苗的叶面积。以玉米幼苗为试验对象对装置性能进行测试,试验结果显示:在装置满载情况下,当相机在X方向和Y方向的移动速度分别为830、32 mm/s时,顶视图模式和侧视图模式下检测装置的平均运行时间分别为190、355 s,检测总耗时为545 s,单株玉米幼苗的平均用时为34 s,相机的平均定位准确率分别为92%和90%,相机定位精度较高;玉米幼苗顶视图、主视图和左视图叶面积与实际叶面积的Pearson相关系数分别为0.901、0.767和0.786,装置检测的玉米幼苗叶面积与实际叶面积相关性强,装置可以满足批量检测玉米幼苗叶面积的需要。
In order to realize the rapid,non-destructive,real-time and high-efficiency detection of corn seedling leaf area,a machine vision-based corn seedling leaf area detection device was designed and built.The detecting device is composed of a frame,a light source device,a jacking rotation system,an image acquisition and analysis system,a detection device control system,and the like.The real-time collection and analysis processing of the top view image and the side view image of the corn seedling were completed by the cooperation of the respective parts,and the corn seedlings leaf area was calculated.The results of the device performance tested with corn seedlings showed that when the device was fully loaded and the moving speed of the camera in the X direction and the Y direction was 830 mm/s and 32 mm/s,respectively.The average running time of the detecting device in the top view mode and the side view mode was 190 s and 355 s,respectively.The total detection time was 545 s,with the average time of single corn seedling of 34 s.The average positioning accuracy of the camera was 92%and 90%,respectively.The positioning accuracy was higher.The Pearson correlation coefficients between the leaf area of the top view,the main view and the left view with the actual leaf area of the corn seedlings were 0.901,0.767 and 0.786,respectively.The leaf area of corn seedlings detected by the device was highly correlated with the actual leaf area.It is indicated that the device can meet the needs of batch detection of the leaf area of corn seedlings.
作者
付豪
万鹏
施家伟
杨万能
FU Hao;WAN Peng;SHI Jiawei;YANG Wanneng(College of Engineering,Huazhong Agricultural University/Key Laboratory of Agricultural Equipment in Mid-Lower Yangtze River,Ministry of Agriculture and Rural Affairs,Wuhan 430070,China;College of Plant Science and Technology,Huazhong Agricultural University,Wuhan 430070,China)
出处
《华中农业大学学报》
CAS
CSCD
北大核心
2020年第1期161-170,共10页
Journal of Huazhong Agricultural University
基金
国家自然科学基金项目（31770397）
关键词
玉米幼苗
叶面积
机器视觉
无损检测
图像采集及分析
作物生长信息
作物幼苗表型性状检测
corn seedlings
leaf area
machine vision
non-destructive testing
image acquisition and analysis
crop growth information
detection phenotypic characteristics of crop seedlings
