植物叶片点云三角剖分与分割的集成算法  

Algorithm of triangulation and segmentation on leaves' point clouds

作　　者：熊威 赵慧洁[1] 李旭东[1] 姜宏志[1] Xiong Wei;Zhao Huijie;Li Xudong;Jiang Hongzhi(Key Laboratory of Precision Opto-mechatronics Technology for Ministry of Education,School of Instrumentation Science ＆ Optoelectronics Engineering,Beihang University,Beijing 100191,China)

机构地区：[1]北京航空航天大学仪器科学与光电工程学院精密光机电一体化技术教育部重点实验,北京100191

出　　处：《计算机应用研究》2018年第8期2524-2526,共3页Application Research of Computers

基　　金：国家自然科学基金资助项目(61227806;61378066)

摘　　要：对植物叶片点云进行三角剖分建立高精度植物叶片模型,对于开展植物冠层结构分析等研究具有重要意义。由三维扫描技术获得的植物叶片点云呈破碎状且空间重叠大、离群点多,传统三角剖分算法的剖分结果不佳。对传统的基于区域生长的三角剖分算法进行改进,在三角剖分的同时对不连续点云进行分割;并对点云内部孔洞进行修复。利用改进后算法对含有大量离群点的植物叶片点云进行三角剖分,结果表明,所提出的算法在完成点云三角剖分和分割的同时剔除了离群点的干扰,提取得到的点云边界及内部孔洞的修复,能满足植物建模中叶片重构的要求;该算法也对不连续物体表面的重构建模提供了有意义的参考。It is significant to research on the structures analysis of the plants’ top level, based on the highly accurate plants’ leave modules set from triangulation of plant leaves’ point clouds. The plant leaves point clouds gotten from 3D scanning technology are scattered, overlapping in space, and with many outliers. The results of the traditional triangulation are not satis- fying . It improved by segmenting non-continual point clouds while generating triangulation. It also recovered the inner holes of point clouds based on triangulation. By this algorithm, it triangulated the plant leaves’ point clouds with many outliers. The result indicates that the proposed algorithm can exclude disturbs of outliers while triangulation and segmentation. Meanwhile, the extracted boundary of point could recovery of inner holes can meet all the requirements for leaves’ restructures in plants’ modeling. It also contributes significantly to the remodeling of non-continual substance’s surfaces.

关 键 词：散乱点云 三角剖分 点云分割 孔洞修复 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]