Cascaded Normal Filtering Neural Network for Geometry-aware Mesh Denoising of Measurement Surfaces

Dingkun ZHU, Yingkui ZHANG, Zhiqi LI, Weiming WANG, Haoran XIE, Mingqiang WEI, Gary CHENG, Fu Lee WANG

Research output: Journal PublicationsJournal Article (refereed)peer-review

Abstract

Mesh denoising is a fundamental component of many disparate reverse engineering applications of measurement surfaces. This article presents a cascaded normal filtering neural network (termed a CNF-Net) for geometry-aware mesh denoising of measurement surfaces. CNF-Net leverages the geometry domain knowledge (GDK) that, a mesh approximates to its underlying surface compactly if all mesh facets at most lie on the surface intersections while not crossing them. Benefiting from the well-estimated underlying geometry of noisy mesh facets, a multiscale guidedly filtered normal descriptor (M-GFND) is formulated, and multiple height maps are constructed from the M-GFND. The height maps can be effectively fed into CNF-Net for learning the transformation matrices between the M-GFND and the ground-truth facet normal. CNF-Net can automatically handle meshes with multiscale geometric features yet corrupted by the noise of different distributions, while existing learning-based wisdoms commonly pursue an overall normal estimation accuracy yet fail to preserve surface significant features. Both visual and numerical evaluations on synthetic and real noise data sets consistently show the clear improvements of CNF-Net over the state-of-the-arts.
Original languageEnglish
Article number2506513
Number of pages14
JournalIEEE Transactions on Instrumentation and Measurement
Volume70
Early online date22 Feb 2021
DOIs
Publication statusPublished - Mar 2021

Bibliographical note

The authors would like to thank the anonymous reviewers for their valuable suggestions. D. Zhu was supported by the Katie Shu Sui Pui Charitable Trust - Academic Publication Fellowship (Reference No.: KSPF2020-02), the Open University of Hong Kong. M. Wei was supported by the National Natural Science Foundation of China (Reference No.: 62032011 and 61502137). W. Wang was supported by the Open University of Hong Kong Research Grant (Reference No. 2020/1.12) and the National Natural Science Foundation of China (Reference No. 61802072). H. Xie was supported by the Direct Grant (Reference No. DR21A5), the Faculty Research Fund (Reference No. DB21A4) and the Lam Woo Research Fund (Reference No. LWI20011) at Lingnan University, Hong Kong. G. Gary was supported by the One-off Special Fund from Central and Faculty Fund in Support of Research from 2019/20 to 2021/22 (Reference No. MIT02/19-20), the Research Cluster Fund (Reference No. RG 78/2019-2020R), the Interdisciplinary Research Scheme of the Dean’s Research Fund 2019-20 (Reference No. FLASS/DRF/IDS-2) of The Education University of Hong Kong.

Publisher Copyright:
IEEE

Copyright:
Copyright 2021 Elsevier B.V., All rights reserved.

Keywords

  • CNF-Net
  • Geometry
  • Geometry domain knowledge
  • Measurement surface
  • Mesh denoising
  • Neural network
  • Neural networks
  • Noise measurement
  • Noise reduction
  • Normal filtering
  • Shape
  • Surface reconstruction
  • Three-dimensional displays

Fingerprint

Dive into the research topics of 'Cascaded Normal Filtering Neural Network for Geometry-aware Mesh Denoising of Measurement Surfaces'. Together they form a unique fingerprint.

Cite this