Author:
Zhu Fan,Li Shaobin,Guo Xiaohan
Abstract
Abstract
3D surface reconstruction with realistic feeling is a hot research field in computer graphics. In order to reconstruct 3D models with high precision, a 3D reconstruction method is proposed in this paper.Based on the BundleFusion reconstruction algorithm, this method uses color images and depth images of indoor scenes obtained by Intel’s realsenseD435i depth camera for rapid 3D reconstruction.An improved ORB feature detection algorithm is used for image feature extraction, and the efficiency of several classical feature detection algorithms in image matching is compared.Experiments show that this method can effectively generate 3d models with higher precision and better sense of reality.
Subject
General Physics and Astronomy
Reference16 articles.
1. Kintinous:Spatially Extended KinectFusion [J];Whelan;Robotics & Autonomous Systems,2012
2. Real-time 3D reconstruction at scale using voxel hashing[J];Nießner;ACM Transactions on Graphics (TOG),2013
3. Very high frame rate volumetric integration of depth images on mobile devices[J];Kahler;IEEE Transactions on Visualization and Computer Graphics,2015
4. ElasticFusion:Dense SLAM Without A Pose Graph[C];Whelan,2015
Cited by
4 articles.
订阅此论文施引文献
订阅此论文施引文献，注册后可以免费订阅5篇论文的施引文献，订阅后可以查看论文全部施引文献