[1]张臻炜,赵 勇,布树辉.一种基于计算机视觉的无人机实时三维重建方法[J].机械与电子,2015,(12):19-22.
　ZHANG Zhenwei,ZHAO Yong BU Shuhui.A UAV Real-time 3D Reconstruction Method Based on Computer Vision[J].Machinery & Electronics,2015,(12):19-22.
点击复制

一种基于计算机视觉的无人机实时三维重建方法
分享到：

《机械与电子》[ISSN:1001-2257/CN:52-1052/TH]

卷:
期数:
2015年12期
页码:
19-22
栏目:
设计与研究
出版日期:
2015-12-25

文章信息/Info

Title:
A UAV Real-time 3D Reconstruction Method Based on Computer Vision
文章编号:
1001-2257(2015)12-0019-04
作者:
张臻炜; 赵 勇; 布树辉
(西北工业大学航空学院,陕西 西安 710072)
Author(s):
ZHANG Zhenwei; ZHAO Yong BU Shuhui
(School of Aeronautics,Northwestern Polytechnical University,Xi'an 710072,China)
关键词:
RGB-D并行定位与建图;  实时三维重建;  图优化
Keywords:
RGB-D SLAM;  3D recon- struction in real-time;  graph optimization
分类号:
TP249
文献标志码:
A
摘要:
针对机器人(或无人机)实时的三维重建中面临的噪声干扰和计算量大的问题,提出一种实时的鲁棒性的同时定位和建图的方法:用RGB-D相机精确估计相机位置的同时重建三维环境。关键思想是测量残差函数——用于估计当前帧的移动量,同时用金字塔模型一步步去逼近真实值。提出一种新的衡量标准用于建立关键帧。然后通过与相邻关键帧的三角化投影给全局地图增加地图点。此外,应用图优化来实现全局优化以获取更高的精度。
Abstract:
The purpose of this paper is to present a robust simultaneous positioning and mapping approach in real-time based on computer vision which shows high efficiency and accuracy to estimate camera poses while mapping the environment with a single RGB-D camera. The key idea is a measurement residual function which is used to estimate motion of the current frame with a coarse to fine scheme. In this paper, we present a novel measurement criterion to determine key frames. Then points of global map are inserted using triangulated projection with neighbor key frames. Moreover we take advantage of the idea of graph optimization to enables global optimization to perform loop-closures and obtain better accuracy.

参考文献/References:

[1] Zhenyou Zhang. A flexible new techniquefor camera calibration[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2000,22(11),1330-1334.
[2] Selig J M.Lie groups and lie algebras in robotics[C]//Computational Noncommutative Algebra and Applications. Springer,2004.
[3] Rosten E,Drummond T.Machine learning for high-speed corner detection[J]. Computer Vision-ECCV2006. Springer,2006.
[4] Kummerle R,Grisetti G,Strasdat H,et.al. G2o: A general framework for graph optimization[C]//IEEE International Conference on Robotics & Automation, 2011:3607-3613.
[5] Endres F,Hess J,Sturm J,et.al. 3-D maping with an RGB-D cameras[J].IEEE Transaction on Robotics,2014, 30(1):177-187.
[6] Guivant J E,Nebot E M. Optimization of the simultaneous localization and map-building algorithm for real-timeimplementation[J]. Robotics & Automation IEEE Transactions on,2001,17(3): 242-257.

备注/Memo

备注/Memo:
收稿日期:2015-09-15
基金项目:2014年西北工业大学本科毕业设计(论文)重点扶持项目 资助
作者简介:张臻炜(1992-),男,福建平和人,硕士研究生,研究方向为计算机图形图像处理、模式识别; 赵 勇(1992-),男,湖南衡阳人,硕士研究生,研究方向为计算机图形图像处理、模式识别。
更新日期/Last Update: 2019-12-25