TreePartNet: Neural Decomposition of Point Clouds for 3D Tree Reconstruction

doi:10.1145/3478513.3480486

CORC > 自动化研究所 > 中国科学院自动化研究所 > 模式识别国家重点实验室 > 三维可视计算

	TreePartNet: Neural Decomposition of Point Clouds for 3D Tree Reconstruction
	Liu, Yanchao 2,3,6; Guo, Jianwei 4,6; Benes, Bedrich 7; Deussen, Oliver 1,5; Zhang, Xiaopeng 4,6; Huang, Hui 2
刊名	ACM TRANSACTIONS ON GRAPHICS
	2021-12-01
卷号	40 期号:6 页码:16
关键词	3D Reconstruction Procedural Modeling Deep Learning Optimization Procedural Generation Geometric Modeling
ISSN号	0730-0301
DOI	10.1145/3478513.3480486
通讯作者	Huang, Hui(hhzhiyan@gmail.com)
英文摘要	We present TreePartNet, a neural network aimed at reconstructing tree geometry from point clouds obtained by scanning real trees. Our key idea is to learn a natural neural decomposition exploiting the assumption that a tree comprises locally cylindrical shapes. In particular, reconstruction is a two-step process. First, two networks are used to detect priors from the point clouds. One detects semantic branching points, and the other network is trained to learn a cylindrical representation of the branches. In the second step, we apply a neural merging module to reduce the cylindrical representation to a final set of generalized cylinders combined by branches. We demonstrate results of reconstructing realistic tree geometry for a variety of input models and with varying input point quality, e.g., noise, outliers, and incompleteness. We evaluate our approach extensively by using data from both synthetic and real trees and comparing it with alternative methods.
资助项目	National Key RD Program[2018YFB2100602] ; NSFC[62172416] ; NSFC[61802406] ; NSFC[U2001206] ; Guangdong Talent Program[2019JC05X328] ; Guangdong Talent Program[00201509] ; Shenzhen Science and Technology Program[RCJC20200714114435012] ; Shenzhen Science and Technology Program[JCYJ20210324120213036] ; Shenzhen Science and Technology Program[JCYJ20180507182222355] ; NSF[10001387] ; Foundation for Food and Agriculture Research[602757] ; Guangdong Laboratory of Artificial Intelligence and Digital Economy (SZ) ; National Engineering Laboratory for Big Data System Computing Technology ; DEGP Key Project[2018KZDXM058]
WOS关键词	SHAPE ; MODELS
WOS研究方向	Computer Science
语种	英语
出版者	ASSOC COMPUTING MACHINERY
WOS记录号	WOS:000729846700037
资助机构	National Key RD Program ; NSFC ; Guangdong Talent Program ; Shenzhen Science and Technology Program ; NSF ; Foundation for Food and Agriculture Research ; Guangdong Laboratory of Artificial Intelligence and Digital Economy (SZ) ; National Engineering Laboratory for Big Data System Computing Technology ; DEGP Key Project
内容类型	期刊论文
源URL	[http://ir.ia.ac.cn/handle/173211/47124]
专题	模式识别国家重点实验室_三维可视计算
通讯作者	Huang, Hui
作者单位	1.Univ Konstanz, Constance, Germany 2.Shenzhen Univ, Shenzhen, Guangdong, Peoples R China 3.CASIA, NLPR, Beijing, Peoples R China 4.Chinese Acad Sci, Inst Automat, NLPR, Beijing, Peoples R China 5.SIAT & Univ Konstanz, Shenzhen, Guangdong, Peoples R China 6.Univ Chinese Acad Sci, Beijing, Peoples R China 7.Purdue Univ, W Lafayette, IN 47907 USA
推荐引用方式 GB/T 7714	Liu, Yanchao,Guo, Jianwei,Benes, Bedrich,et al. TreePartNet: Neural Decomposition of Point Clouds for 3D Tree Reconstruction[J]. ACM TRANSACTIONS ON GRAPHICS,2021,40(6):16.
APA	Liu, Yanchao,Guo, Jianwei,Benes, Bedrich,Deussen, Oliver,Zhang, Xiaopeng,&Huang, Hui.(2021).TreePartNet: Neural Decomposition of Point Clouds for 3D Tree Reconstruction.ACM TRANSACTIONS ON GRAPHICS,40(6),16.
MLA	Liu, Yanchao,et al."TreePartNet: Neural Decomposition of Point Clouds for 3D Tree Reconstruction".ACM TRANSACTIONS ON GRAPHICS 40.6(2021):16.