基于三维激光点云的靶标探测系统研究与试验  

Research and test of target detection system based on 3D laser point cloud

作　　者：齐亚聪 陈毅飞 杨会民 王学农[1,2,3] QI Yacong;CHEN Yifei;YANG Huimin;WANG Xuenong(College of Mechanical and Electrical Engineering,Xinjiang Agricultural University,Urumqi 830052,China;Scientific Observation and Experimental Station of Forest Fruit,Cotton and Facility,Agriculture Equipment,Ministry of Agriculture and Rural Affairs,Urumqi 830091,China;Agricultural Mechanization Institute,Xinjiang Academy of Agricultural Sciences,Urumqi 830091,China)

机构地区：[1]新疆农业大学机电工程学院,乌鲁木齐830052 [2]农业农村部林果棉装备科学观测实验站,乌鲁木齐830091 [3]新疆农业科学院农业机械化研究所,乌鲁木齐830091

出　　处：《新疆农业科学》2023年第11期2833-2841,共9页Xinjiang Agricultural Sciences

基　　金：新疆设施农业智能管控技术重点实验室(XJYS1703);新疆农业科学院青年科技骨干创新能力培养项目“基于2D与3D数据多模态融合的植株冠层靶标识别与分割方法研究”(xjnkg-2023026);教育部第二批新工科研究与实践项目“新工科教育体系下农业工程类专业人才培养质量提升路径探索与实践”(E-SPNL20202326,地方高校组)。

摘　　要：【目的】使用R-Fans-32三维激光雷达(LiDAR)研究植株三维激光点云与植株叶面积之间的关系,为变量喷雾系统提供数据支撑。【方法】假设植株激光点云数量与叶面积之间存在线性关系。搭建基于三维激光点云的靶标探测的试验系统,先测量靶标植株的高度来探究该探测系统的精度,激光雷达以10Hz的扫描频率和1m的探测距离实现对10株番茄的三维点云数据的获取,激光雷达上位机软件Ctrlview实现对三维激光点云数据的储存。利用Cloud Compare软件对储存的点云数据进行处理,利用LiDAR360软件对植株进行高度测量和点云数量的获取。对采集的植株点云进行数量统计,利用CL-202植物叶面积测量仪对采摘的靶标植株叶片测量叶面积,验证植株点云与叶面积之间的关系。【结果】激光雷达探测所得到的番茄植株的高度与手工测量值的最大相对误差为7.92%。利用线性函数拟合植株点云数量与叶面积,拟合度为0.7805,最大相对误差为5.64%。【结论】设计了一种用于探究基于激光点云的变量喷雾系统可行性的试验系统,依据三维激光点云计算植株的叶面积精度良好,R-Fans-32三维激光雷达可作为变量喷雾系统的探测部件。【Objective】Variable spray system can reduce the waste of liquid medicine and reduce the pollution of liquid medicine to land and water.The acquisition of plant geometric parameter information is an important prerequisite for implementing variable spray.R-Fans-32 three-dimensional laser radar(LiDAR)was used to explore the relationship between plant 3D laser point cloud and plant leaf area,providing data support for variable spray machine.【Methods】It was assumed that there was a linear relationship between the number of laser point cloud and leaf area.The target detection based on 3 d laser point cloud of the test system was set to measure the height of the target plant to explore the accuracy of laser radar,laser radar with 10 Hz scanning frequency and 1m of the detection range of 10 strains of tomato three-dimensional point cloud data acquisition of laser radar PC software Ctrlview implementation of 3 d laser point cloud data storage.Cloud Compare software was used to process the stored point cloud data,and LiDAR360 software was used to measure the height of plants and obtain the number of point Cloud.The quantity of collected plant point cloud was counted,and the leaf area of picked target plant leaves was measured by CL-202 plant leaf area meter to verify the relationship between plant point cloud and leaf area.【Results】The experimental results showed that the maximum relative error between the height of tomato plants detected by lidar and manual measurement was 7.92%.The linear function was used to fit the number of plant point cloud and leaf area,with a fitting degree of 0.7805 and a maximum relative error of 5.64%.【Conclusion】An experimental system was designed to explore the feasibility of variable spray system based on laser point cloud.The accuracy of plant leaf area prediction based on 3d laser point cloud was good,and 3D LIDAR R-FAN-32 could be used as the component of variable spray system for crop detection.

关 键 词：变量喷雾 激光雷达 信息采集 三维点云 叶面积 测量 

分 类 号：S12[农业科学—农业基础科学]