基于三维激光点云的靶标叶面积密度计算方法

发布时间：2018-03-15 00:14

  本文选题：变量喷雾　切入点：激光雷达扫描　出处：《农业机械学报》2017年11期 　论文类型：期刊论文

【摘要】：为向变量喷雾系统施药量的计算提供数据基础,提出了靶标喷施区域叶面积密度参数的计算方法。靶标三维点云数据由二维激光雷达传感器沿果树行直线运动间接获取。在假设各喷施子区域内叶片面积变化相对较小的条件下,基于Matlab曲线拟合工具箱cftool分析并验证了点云数与叶片数之间存在函数关系。曲线拟合结果表明,利用高斯函数、多项式函数与指数函数拟合点云数与叶片数,决定系数分别为0.925 7、0.931 0与0.936 4,指数函数拟合效果最好。相对误差分析结果表明,基于3种拟合函数,枝叶茂密区域相对误差最小为11.46%,枝叶中等茂密区域相对误差最小为11.05%,枝叶稀疏区域相对误差最小为35.50%。基于确定的点云数与叶片数间的函数方程,经系数变换后可计算出叶面积密度参数。
[Abstract]:To provide a data basis for the calculation of the amount of pesticide applied to a variable spray system, A method for calculating the leaf area density parameters of target sprayed area is presented. The 3D point cloud data of target are obtained indirectly by two-dimensional lidar sensor moving in a straight line along the fruit tree. The change phase of leaf area is assumed to be within each sprayed sub-region. For smaller conditions, Based on Matlab curve fitting toolbox cftool, the function relationship between point cloud number and blade number is analyzed and verified. The curve fitting results show that point cloud number and leaf number are fitted by Gao Si function, polynomial function and exponential function. The determination coefficients are 0.925 7 / 0.931 0 and 0.936 4, respectively. The results of relative error analysis show that, based on the three fitting functions, the exponential function is the best. The minimum relative error in the dense region of branches and leaves is 11.46%, the minimum relative error in the middle dense region of branches and leaves is 11.05, and the minimum relative error in sparse region of branches and leaves is 35.50.Based on the function equation between the number of point clouds and the number of leaves, The parameters of leaf area density can be calculated by coefficient transformation.
【作者单位】： 江苏省农业科学院农业设施与装备研究所;南京农业大学工学院;江苏省农业科学院园艺所;
【基金】：国家梨产业技术体系专项(CARS-29-18) 江苏省农业科技自主创新资金项目(CX(15)1023) 江苏省农业科学院基本科研业务费专项(ZX(16)3006)
【分类号】：S49;TN958.98
，

本文编号：1613571