基于激光测距的橡胶树树干轮廓曲线自动化提取方法  

Automatic extraction method of the outline curve of rubber treebased on the laser-ranging technology

作　　者：黎土煜 刘青 郑勇[1] 黄敞[1] 曹建华[1] 王玲玲 Li Tuyu;Liu Qing;Zheng Yong;Huang Chang;Cao Jianhua;Wang Lingling(Rubber Research Institute,Chinese Academy of Tropical Agricultural Sciences,Haikou,571101,China;Zhanjiang Agricultural Reclamation Modern Agriculture Development Co.,Ltd.,Zhanjiang,524200,China)

机构地区：[1]中国热带农业科学院橡胶研究所,海口市571101 [2]广东省湛江农垦现代农业发展有限公司,广东湛江524200

出　　处：《中国农机化学报》2024年第2期194-199,F0002,共7页Journal of Chinese Agricultural Mechanization

基　　金：中央级科研事业单位基本科研业务费专项(1630022022005);海南省自然科学基金项目(320QN349、323QN273)。

摘　　要：为获取橡胶树树干轮廓曲线模型,解决割胶装备难以精准贴合树干轮廓导致作业精度差、仿形效果不理想的关键问题,提出一种橡胶树树干轮廓曲线自动化提取方法。研制橡胶树树干轮廓点云数据自动采集平台,该平台由电源模块、控制模块、驱动模块、采集模块、数据传输模块以及数据记录处理模块组成,可实现树干外轮廓曲线点云数据的自动采集、传输、记录与处理。开展橡胶树树干轮廓曲线自动化提取试验,基于统计学方差分析法,通过曲线面积对比分析树干轮廓提取精度。结果表明,连续采集30组橡胶树树干轮廓点云数据,其轮廓特征有效识别度较高,轮廓特征点位置与实际位置基本符合;轮廓曲线重复提取变异系数为0.001 3;树干轮廓曲线提取精度不受树干位置和采集方向变化的影响,该技术方案具有较强可行性。In order to acquire the contour curve representation of rubber tree trunks and address the significant industrial challenge stemming from inadequate operational precision and suboptimal shaping outcomes resulting from the challenges associated with accurately fitting tapping equipment to the trunk contour,a method to automatically extract rubber tree outline curve based on laser-ranging technology was proposed.The automatic data acquisition platform of rubber tree trunk contour point cloud is developed,which is composed of power module,control module,drive module,acquisition module,data transmission module,data recording module and processing module.The platform has the function of automatic collection,transmission,recording and processing of the outline curve point cloud data.The automatic extraction experiment of rubber tree outline curve was carried out,and the accuracy of rubber tree outline extraction was analyzed by comparing the outline curve area based on statistical variance analysis.The results showed that the position of contour feature points was basically consistent with the actual position with a high recognition when 30 groups of contour point cloud data were collected continuously,and the coefficient of variation of repeated extraction of contour curves was 0.0013.The extraction accuracy of trunk contour curve is not affected by the change of trunk position and acquisition direction,which means that the technical scheme is very feasible.

关 键 词：橡胶树 激光测距 点云数据 轮廓曲线 自动提取 

分 类 号：S794.1[农业科学—林木遗传育种]