由于现阶段对于花朵三维点云的特征提取研究还不充足,以及三维点云数据的无序性,文章旨在提出一种基于圆柱坐标系的花朵点云特征提取方法,以单个百合花点云数据为研究对象,利用圆柱坐标系的半径、角度和高度等参数来建立与点云中点之间的拓扑关系,从而更好地进行点云特征提取。经实验分析,该方法能有效提取出百合花朵点云的6个花瓣和1个花心特征,极大程度上为后续点云分割奠定基础。
Abstract
Due to the insufficient research on feature extraction of 3D point clouds of flowers at the current stage and the disorder of 3D point cloud data, this paper aims to propose a flower point cloud feature extraction method based on the cylindrical coordinate system. Taking the point cloud data of a single lily flower as the research object, the method uses parameters such as radius, angle and height of the cylindrical coordinate system to establish the topological relationship between points in the point cloud, so as to achieve better point cloud feature extraction. Experimental analysis shows that this method can effectively extract the features of 6 petals from the lily flower point cloud, which largely lays a foundation for subsequent point cloud segmentation.
关键词
三维点云 /
特征提取 /
圆柱坐标系 /
花朵
Key words
3D point cloud /
feature extraction /
cylindrical coordinate system /
flower
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] 唐家阳. 基于深度学习的叶菜类植物点云分割方法研究与实现[D].杨凌:西北农林科技大学,2024.
[2] 赖亦斌. 基于三维点云的小型植物表型提取研究[D].桂林:广西师范大学,2021.
[3] 蔡东娜. 重瓣花卉植物的三维重建与动态模拟研究[D].北京:北京林业大学,2018.
[4] 彭丽君. 基于点云的植物果实三维数据精细测量与品质分级[D].抚州:东华理工大学,2021.
[5] 孙智慧. 基于点云数据的植物形态建模方法研究[D].北京:首都师范大学,2012.
[6] 张越,翟福琪,蔡孙宝,等. 基于点云数据的植物叶片特征提取及三维重建[J].中国测试,2021,47(8):6-12.
[7] 柯映林,范树迁.基于点云的边界特征直接提取技术[J].机械工程学报,2004(9):116-120.
[8] 舒丽君. 基于三维点云树枝特征提取及重建[D].抚州:东华理工大学,2021.