摘要
为及时准确地获取棉花出苗信息,首先,利用无人机搭载数字成像设备获取棉花出苗期图像,通过超绿算法和手动阈值法分割出棉苗;然后,应用图像去相关拉伸和K均值聚类算法实现对地膜区域的分割,分析地膜垂直投影统计图,准确获得棉苗行数和地膜规划区域,实现对地膜规划区域外杂草的去除和棉苗区域的分组;最后,应用最小二乘法对棉苗进行直线拟合,实现了苗行的精确检测和单条苗行直线距离。通过棉苗位置距离中位数获得株距,从而得到棉花播种数。结果显示,出苗行数检测准确,误差为0,棉花出苗率检测准确率最高为90.45%,棉花播种数误差仅为5%。所提出的方法实现了大面积棉田出苗信息的快速提取,为棉苗管理的精细农业提供技术支持,为田间苗行检测提供新方法。
In order to obtain cotton seeding emergence information timely and accurately,a digital imaging equipment is mounted on UAV(unmanned aerial vehicle) to get the images of cotton seeding emergence period. The cotton seedlings are segmented by extra-green algorithm and manual threshold value method. The image decorrelation stretching and K-means clustering algorithm are used to segment the plastic mulch area. The mulch vertical projection statistical chart is analyzed to accurately obtain the row number of cotton seedlings and plastic mulch planning area,so as to remove the weeds outside the plastic mulch planning area and group the cotton seedling area. The least square method is used to implement linear fitting of cotton seedlings,and accurately detect the seedling rows and the linear distance of single seedling row. The plant spacing is obtained by the median distance of cotton seedling position,so as to get the number of cotton sowing. The results show that the method′ s detection precision of seedling row number is accurate(the error is 0),the highest accuracy rate of cotton seedling emergence rate is 90.45%,and the error of cotton sowing number is only 5%. Therefore,the proposed method can realize the rapid extraction of seeding emergence information in large area of cotton field. It provides technical support for precision agriculture of seedling management and offers a new method for seedling line detection in the field.
作者
朱松松
陈至坤
张怡
ZHU Songsong;CHEN Zhikun;ZHANG Yi(College of Electrical Engineering,North China University of Science and Technology,Tangshan 063210,China)
出处
《现代电子技术》
2022年第1期61-64,共4页
Modern Electronics Technique
基金
国家自然科学基金项目(61803154)。
关键词
棉花出苗信息
无人机视觉系统
地膜区域分割
株距获取
苗行检测
棉花播种数
cotton seeding emergence information
UAV visual system
mulch area segmentation
plant spacing acquisition
seedling row detection
cotton seeding number
