General Point Cloud Feature Extraction and Matching Algorithem

3D scene perception and reconstruction based on laser point cloud is an important research direction in digital city, virtual reality and other fields. The common point cloud matching method generally refers to solving the point cloud correspondence under low dynamic conditions, but it cannot solve...

Full description

Saved in:
Bibliographic Details
Published in2022 19th International Computer Conference on Wavelet Active Media Technology and Information Processing (ICCWAMTIP) pp. 1 - 6
Main Authors Jingxi, Gao, Feng, Gao, Shuhua, Teng, Xing, Zhang, Zhiguo, Tan, Ling, Wang
Format Conference Proceeding
LanguageEnglish
Published IEEE 16.12.2022
Subjects
Feature matching
Heuristic algorithms
High precision
Information processing
Large dynamic
Lasers
Point cloud compression
Point cloud feature extraction
Three-dimensional displays
Urban areas
Virtual reality
Online AccessGet full text

Cover

More Information
Summary:3D scene perception and reconstruction based on laser point cloud is an important research direction in digital city, virtual reality and other fields. The common point cloud matching method generally refers to solving the point cloud correspondence under low dynamic conditions, but it cannot solve the point cloud matching problem under large dynamic conditions. Therefore, this paper proposes a new point cloud feature extraction and matching method, which is compatible with point cloud matching under low dynamic and large dynamic conditions. Experiments show that the proposed algorithm can adaptively solve the point cloud matching problem under various dynamic conditions, and has high accuracy.
ISBN:9781665493871
1665493879
ISSN:2576-8964
DOI:10.1109/ICCWAMTIP56608.2022.10016493