Robust point cloud lightweighting with multi-scale adaptive filtering and entropy-driven subdivision

Weibo Zeng, Xinyu Gao, Qi Lu, Ning Zhu, Mengchan Li, Wenjing Cai · 2026

Read Robust point cloud lightweighting with multi-scale adaptive filtering and entropy-driven subdivision by Weibo Zeng, Xinyu Gao, Qi Lu, Ning Zhu, Mengchan Li, Wenjing Cai (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/plos-pone-0353953.txt.

Description

Zeng and colleagues, in a 2026 paper, present a better way to slim down point clouds, the millions of dots that scanners collect to map real scenes. Slimming cuts file size and computing load, but crude cuts either leave noise behind or erase sharp edges. Their method works at several scales at once.

Three ideas combine. A sliding filter smooths noise while guarding sharp features. Space is split into boxes guided by how much information each area holds, so dense areas keep detail and sparse areas waste less memory. A strengthened skeleton method then extracts the core shape of objects with fewer breaks. On their own scans of buildings, plants, and roads, and on a public test set, the method removed most noise. It kept edges well, and rebuilt surfaces with the lowest errors among the methods tried. The authors aim it at city modeling and scene rebuilding.

Catalog record

CREATORWeibo Zeng, Xinyu Gao, Qi Lu, Ning Zhu, Mengchan Li, Wenjing Cai
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE91459
LICENSECC_BY
RIGHTSCC BY
SOURCEPLOS

Source: https://doi.org/10.1371/journal.pone.0353953

Browse the full library · About the Tower of Babel