Fast surface reconstruction algorithm with adaptive step size.

In (Dai et al. 2023), the authors proposed a fast algorithm for surface reconstruction that converges rapidly from point cloud data by alternating Anderson extrapolation with implicit progressive iterative approximation (I-PIA). This algorithm employs a fixed step size during iterations to enhance c...

Full description

Saved in:
Bibliographic Details
Main Authors: Jingguo Dai, Yeqing Yi, Chengzhi Liu
Format: Article
Language:English
Published: Public Library of Science (PLoS) 2025-01-01
Series:PLoS ONE
Online Access:https://doi.org/10.1371/journal.pone.0314756
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1832540168058306560
author Jingguo Dai
Yeqing Yi
Chengzhi Liu
author_facet Jingguo Dai
Yeqing Yi
Chengzhi Liu
author_sort Jingguo Dai
collection DOAJ
description In (Dai et al. 2023), the authors proposed a fast algorithm for surface reconstruction that converges rapidly from point cloud data by alternating Anderson extrapolation with implicit progressive iterative approximation (I-PIA). This algorithm employs a fixed step size during iterations to enhance convergence. To further improve the computational efficiency, an adaptive step size adjustment strategy for surface reconstruction algorithm is investigated. During each iteration, the step size is adaptively chosen based on the current residual-larger residuals may necessitate larger steps, while smaller ones might permit smaller steps. Numerical experiments indicate that, for equivalent reconstruction errors, the adaptive step size algorithm demands substantially fewer iterations and less computation time than the fixed step size approach. These improvements robustly enhance computational performance in surface reconstruction, offering valuable insights for further research and applications.
format Article
id doaj-art-333e575110a44beda2339ec48f5e4f03
institution Kabale University
issn 1932-6203
language English
publishDate 2025-01-01
publisher Public Library of Science (PLoS)
record_format Article
series PLoS ONE
spelling doaj-art-333e575110a44beda2339ec48f5e4f032025-02-05T05:32:05ZengPublic Library of Science (PLoS)PLoS ONE1932-62032025-01-01201e031475610.1371/journal.pone.0314756Fast surface reconstruction algorithm with adaptive step size.Jingguo DaiYeqing YiChengzhi LiuIn (Dai et al. 2023), the authors proposed a fast algorithm for surface reconstruction that converges rapidly from point cloud data by alternating Anderson extrapolation with implicit progressive iterative approximation (I-PIA). This algorithm employs a fixed step size during iterations to enhance convergence. To further improve the computational efficiency, an adaptive step size adjustment strategy for surface reconstruction algorithm is investigated. During each iteration, the step size is adaptively chosen based on the current residual-larger residuals may necessitate larger steps, while smaller ones might permit smaller steps. Numerical experiments indicate that, for equivalent reconstruction errors, the adaptive step size algorithm demands substantially fewer iterations and less computation time than the fixed step size approach. These improvements robustly enhance computational performance in surface reconstruction, offering valuable insights for further research and applications.https://doi.org/10.1371/journal.pone.0314756
spellingShingle Jingguo Dai
Yeqing Yi
Chengzhi Liu
Fast surface reconstruction algorithm with adaptive step size.
PLoS ONE
title Fast surface reconstruction algorithm with adaptive step size.
title_full Fast surface reconstruction algorithm with adaptive step size.
title_fullStr Fast surface reconstruction algorithm with adaptive step size.
title_full_unstemmed Fast surface reconstruction algorithm with adaptive step size.
title_short Fast surface reconstruction algorithm with adaptive step size.
title_sort fast surface reconstruction algorithm with adaptive step size
url https://doi.org/10.1371/journal.pone.0314756
work_keys_str_mv AT jingguodai fastsurfacereconstructionalgorithmwithadaptivestepsize
AT yeqingyi fastsurfacereconstructionalgorithmwithadaptivestepsize
AT chengzhiliu fastsurfacereconstructionalgorithmwithadaptivestepsize