面向多角度视图三维重建的基础矩阵求解

冯宏伟; 周亚培; 冯筠; 王红玉; 刘晓宁

doi:10.3788/OPE.20162413.0567

您当前的位置：

首页 >

文章列表页 >

面向多角度视图三维重建的基础矩阵求解

更新时间：2020-08-13

- 面向多角度视图三维重建的基础矩阵求解
- Fundamental matrix estimation for 3D reconstruction towards multi-perspective views
- 光学精密工程 2016年24卷第10s期页码：567-574
- 作者机构：
  
  西北大学信息科学与技术学院,陕西西安,710127
- 作者简介：
- 基金信息：
  
  国家自然科学基金资助项目（No.61305032();No.61372046）()
- DOI：10.3788/OPE.20162413.0567
  中图分类号： TP391
- 收稿日期：2016-06-05，
  
  修回日期：2016-06-20，
  
  纸质出版日期：2016-11-14
- 稿件说明：
移动端阅览
冯宏伟, 周亚培, 冯筠等. 面向多角度视图三维重建的基础矩阵求解[J]. 光学精密工程, 2016,24(10s): 567-574

FENG Hong-wei, ZHOU Ya-pei, FENG Jun etc. Fundamental matrix estimation for 3D reconstruction towards multi-perspective views[J]. Editorial Office of Optics and Precision Engineering, 2016,24(10s): 567-574
冯宏伟, 周亚培, 冯筠等. 面向多角度视图三维重建的基础矩阵求解[J]. 光学精密工程, 2016,24(10s): 567-574 DOI： 10.3788/OPE.20162413.0567.

FENG Hong-wei, ZHOU Ya-pei, FENG Jun etc. Fundamental matrix estimation for 3D reconstruction towards multi-perspective views[J]. Editorial Office of Optics and Precision Engineering, 2016,24(10s): 567-574 DOI： 10.3788/OPE.20162413.0567.

摘要

针对已有的求解基础矩阵算法求解精度不高，稳定性不理想的问题，提出了一种面向多角度视图三维重建的基础矩阵求解算法。首先提出了基于模拟退火的抽样策略（Simulated Aneal Sampling，SAS），把匹配特征点集划分成不同类别，通过在每个类别中以不同概率接受抽取的特征点对的形式完成抽样；然后提出了一种基于匹配点对的重投影误差内点筛选算法，并设计了基于SAS的基础矩阵求解算法。实验结果表明该算法的精确性和稳定性比目前流行的两种算法提高了10倍左右，能够有效地提高三维重建的准确率。

Abstract

In consideration of the relatively low precision and stability in the existing methods for Fundamental Matrix Estimation(FME)

a FME method for 3D reconstruction of multi-perspective views is put forward. Firstly

sampling strategy based on Simulated Aneal Sampling (SAS) is proposed. Specifically

the candidate feature points are divided into different categories

and the sampling is fulfilled by extracting the feature point pair from each category in different probability; Furthermore

an interior points filtering algorithm is proposed based on re-projection error of matched points pair; Finally

a FME algorithm based on SAS is designed. It indicates that the precision and stability achieved by this solution is approximately about 10 times higher than that of state-of-art algorithms

and such solution can effectively improve the precision rate of three-dimensional reconstruction.

关键词

Keywords

references

陈杰春, 丁振良, 袁峰. 三维重构不确定度的两阶估计方法[J]. 光学精密工程, 2008, 16(6):1110-1116. CHEN J CH DING ZH L YUAN F.Two-stage uncertainty evaluation of 3D reconstruction[J].Opt. Precision Eng., 2008, 16(6):1110-1116. (in Chinese)

王欣, 袁坤, 于晓,等. 基于运动恢复的双目视觉三维重建系统设计[J]. 光学精密工程, 2014, 22(5):1379-1387. WAND X,YUAN K,YU X,et al.Design of binocular vision 3D reconstruction system based on motion recovery[J].Opt. Precision Eng., 2014, 22(5):1379-1387. (in Chinese)

潘淑杰, 董守平, 王红丽. 应用灭点标定的立体视觉自由曲面三维重建[J]. 光学精密工程, 2009, 17(5):1127-1133. PAN SH J, DONG SH P, WANG H L.Three dimensional reconstruction of stereo vision free-form surface based on vanishing-point calibration[J].Opt. Precision Eng., 2009, 17(5):1127-1133. (in Chinese)

ZHOU C, GUNEY F, WANG Y, et al.. Exploiting object similarity in 3D reconstruction[C]. Proceedings of the IEEE International Conference on Computer Vision, 2015:2201-2209.

ESPUNY F, MONASSE P. Singular Vector Methods for Fundamental Matrix Computation[M]. Berlin:Image and Video Technology. Springer Berlin Heidelberg, 2014:290-301.

ZHENG Y, SUGIMOTO S, OKUTOMI M. A practical rank-constrained eight-point algorithm for fundamental matrix estimation[C]. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2013:1546-1553.

陈付幸, 王润生. 基础矩阵估计的聚类分析算法[J]. 计算机辅助设计与图形学学报, 2005(10):2251-2256. CHEN F X,WANG R S.Clustering algorithm for the fundamental matrix estimation[J]Journal of Computer-Aided Design & Computer Graphics, 2005(10):2251-2256. (in Chinese)

ZHANG M, WANG G, CHAO H, et al.. Fast and Robust Algorithm for Fundamental Matrix Estimation[M]. Image Analysis and Recognition. Springer International Publishing, 2015:316-322.

ZHOU F, ZHONG C, ZHENG Q. Method for fundamental matrix estimation combined with feature lines[J]. Neurocomputing, 2015, 160:300-307.

周凡, 邵世维, 吴建华,等. 一种联合直线特征的基础矩阵计算方法[J]. 光学学报, 2013(10):188-195. ZHOU F,SHAO SH W,Wu J H,et al..Method for fundamental matrix estimation combined with line features[J].Acta Optica Sinica, 2013(10):188-195. (in Chinese)

YANG J L, LI H D, JIA Y D. Optimal essential matrix estimation via inlier-set maximization[C]. European Conference on Computer Vision. 2014:111-126.

MACHh C A C. Random Sample Consensus:a paradigm for model fitting with application to image analysis and automated cartography[C]. 1981.

BELLAVIA F, TEGOLO D. noRANSAC for fundamental matrix estimation[C]. Proc. of he British Machine Vision Conference, 2011:1-98.

NI K, JIN H, DELLAERT F. Groupsac:Efficient consensus in the presence of groupings[C]. 2009 IEEE 12th International Conference on Computer Vision, IEEE, 2009:2193-2200.

甄艳, 刘学军, 王美珍. 匹配点分布密度约束下的基础矩阵估计[J]. 武汉大学学报(信息科学版), 2013, 38(10):1167-1171. ZHEN Y,LIU X J,WANG M ZH. Fundamental matrix estimation based on inlier distributions constraint[J]. Geomatics and Information Science of Wuhan University, 2013, 38(10):1167-1171. (in Chinese)

ZHU Q, LIU C, CAI C. A novel robot visual homing method based on SIFT features[J]. Sensors, 2015, 15(10):26063-26084. (in Chinese)

张从鹏, 魏学光. 基于Harris角点的矩形检测[J]. 光学精密工程, 2014, 22(8):2259-2266. ZHANU C P,WEI X G.Rectangle detection based on Harris corner[J]. Opt. Precision Eng., 2014, 22(8):2259-2266. (in Chinese)

DHAMECHA T I, SHARMA P, SINGH R, et al.. On effectiveness of histogram of oriented gradient features for visible to near infrared face matching[C]. 201422nd International Conference on Pattern Recognition(ICPR), 2014:1788-1793.

WANG GH, ZELEK J S, WU Q M J, et al.. Robust rank-4 affine factorization for structure from motion[C]. 2013 IEEE Workshop on Applications of Computer Vision (WACV), 2013:180-185.

http://www.robots.ox.ac.uk/~vgg/data1.html

浏览量

307

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

基于平面补丁的自适应八叉树三维图像重建

基于深度相机的大场景三维重建

基于激光扫描的密集泡状流三维重建与优化

基于关系图的无人机影像三维重建

用于弱纹理场景三维重建的机器人视觉系统