摘要
鉴于多视图三维重建算法的特征检测和匹配能力较差,文章提出一种基于改进ORB特征提取的快速三维重建算法。该算法针对传统算法在建筑数据采集中的重建效果差和运行效率低的问题进行了优化,首先运用改进的ORB算法进行特征点提取,采用GMS算法进行特征匹配,得到匹配结果后利用全局式SFM对场景相机位姿进行估计,之后再进行光束法平差,最后利用PMVS算法得到稠密重建的结果。实验结果表明,该算法在三维重建运行效率上有所提升,能够满足快速的三维重建要求。
In view of the poor feature detection and matching ability of multi view iterative reconstruction algorithm,this paper proposes a fast iterative reconstruction algorithm based on improved ORB feature extraction.This algorithm has been optimized to address the issues of poor reconstruction performance and low operational efficiency of traditional algorithms in building data collection.Firstly,the improved ORB algorithm is used for feature point extraction,and the GMS algorithm is used for feature matching.After obtaining the matching results,the global SFM is used to estimate the camera pose of the scene,followed by bundle adjustment.Finally,the PMVS algorithm is used to obtain the dense reconstruction results.The experimental results show that the algorithm improves the running efficiency of iterative reconstruction and can meet the requirements of fast iterative reconstruction.
作者
罗庆
唐清善
LUO Qing;TANG Qingshan(School of Physics&Electronic Science,Changsha University of Science&Technology,Changsha 410114,China)
出处
《现代信息科技》
2023年第15期83-86,90,共5页
Modern Information Technology
基金
湖南省研究生科研创新项目(CX2207)
湖南省水利厅基于图像识别水位监测系统的研究项目(XSKJ2021000-27)。
