改进拟牛顿算法的点云稠密化应用研究Research on application of improved Quasi-Newton algorithm in point cloud densification
黎华,凯吾沙·塔依尔,林木森,蒲睿,吴浩
摘要(Abstract)：
针对在三维重建中,利用图像序列重建的方法耗时过长、模型精度低的问题,该文提出一种对离散点云稠密化的优化方法。以图像序列的三维模型重建为背景,在增量式运动恢复结构算法(SfM)生成三维稀疏点云的基础上,对目标物体表面的稀疏离散点进行稠密化,并提出了一种针对基于面片的多视角立体几何算法(PMVS)时间复杂度的优化方法。针对稠密点云重建过程中PMVS算法耗时过长的问题,构建一种基于原始的拟牛顿优化方法的改进算法。在PMVS算法的面片优化部分中,对拟牛顿算法(BFGS)的修正矩阵迭代公式进行改进,确保全局收敛的同时也提高了收敛速度。实验结果表明,该文使用的改进算法不仅加快了三维重建的速度,而且适当提升了稠密点云重建的质量。
关键词(KeyWords)： 三维重建;运动恢复结构;PMVS算法;改进的拟牛顿算法
基金项目(Foundation): 国家自然科学基金项目(42071358,41301588);; 湖北省重点实验室(三峡大学)开放基金项目(2016KJZ05)
作者(Author): 黎华,凯吾沙·塔依尔,林木森,蒲睿,吴浩
DOI: 10.16251/j.cnki.1009-2307.2021.12.012
参考文献(References)：
- [1] LONGUET-HIGGINS H C.A computer algorithm for reconstructing a scene from two projections[J].Nature,1981,293(5828):133-135.
- [2] FRUH C,ZAKHOR A.3D model generation for cities using aerial photographs and ground level laser scans[C]//Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition,CVPR 2001.Kauai,HI,USA:IEEE,2001:II.
- [3] STAMOS I,ALIEN P K.Automatic registration of 2-D with 3-D imagery in urban environments[C]//Proceedings Eighth IEEE International Conference on Computer Vision,ICCV 2001.Vancouver,BC,Canada:IEEE,2001:731-736.
- [4] TAHRI O,BOUTAT D,MEZOUAR Y.Brunovsky’s linear form of incremental structure from motion[J].IEEE Transactions on Robotics,2017,33(6):1491-1499.
- [5] TRAN S,DAVIS L.3D surface reconstruction using graph cuts with surface constraints[M]//Computer Vision-ECCV 2006.Berlin,Heidelberg:Springer Berlin Heidelberg,2006:219-231.
- [6] FURUKAWA Y,PONCE J.Carved visual hulls for image-based modeling[J].International Journal of Computer Vision,2009,81(1):53-67.
- [7] FURUKAWA Y,PONCE J.Accurate,dense,and robust multiview stereopsis[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2010,32(8):1362-1376.
- [8] H?NE C,ZACH C,COHEN A,et al.Joint 3D scene reconstruction and class segmentation[C]//2013 IEEE Conference on Computer Vision and Pattern Recognition.Portland,OR,USA:IEEE,2013:97-104.
- [9] LI Bo,VENKATESH Y V,KASSIM A,et al.Improving PMVS algorithm for 3D scene reconstruction from sparse stereo pairs[M]//Lecture Notes in Computer Science.Cham:Springer International Publishing,2013:221-232.
- [10] 杨文博，孙博文.双约束条件下PMVS的改进算法[J].计算机科学与探索，2017,11(11):1804-1815.(YANG Wenbo,SUN Bowen.Improved PMVS algorithm with double constraints[J].Journal of Frontiers of Computer Science and Technology,2017,11(11):1804-1815.)
- [11] 史利民，郭复胜，胡占义.利用空间几何信息的改进PMVS算法[J].自动化学报，2011,37(5):560-568.(SHI Limin,GUO Fusheng,HU Zhanyi.An improved PMVS through scene geometric information[J].Acta Automatica Sinica,2011,37(5):560-568.)
- [12] 陈冉.基于准稠稠密匹配配方法的的PMVS算算法改进[D].北京：北京工业大学，2014.(CHEN Ran.Improving PMVS based on quasi-dense matching[D].Beijing:Beijing University of Technology,2014.)
- [13] 倪标，王铮尧.一种倾斜影像密集匹配算法的设计与实现[J].地理空间信息，2020,18(7):66-69.(NI Biao,WANG Zhengyao.Design and implementation of dense matching algorithm based on oblique images[J].Geospatial Information,2020,18(7):66-69.)
- [14] 隗娜，郭向坤，董志勇，等.基于像素插值的改进PMVS稠密重建方法[J].计算机系统应用，2019,28(7):157-161.(KUI Na,GUO Xiangkun,DONG Zhiyong,et al.Dense reconstruction method based on improved PMVS and pixel interpolation[J].Computer Systems & Applications,2019,28(7):157-161.)
- [15] 皇甫瑞.一类新拟牛顿算法及其收敛性[D].临汾：山西师范大学，2010.(HUANGPU Rui.A new class of quasi-Newton algorithm and its convergence[D].Linfen:Shanxi Normal University,2010.)
- [16] MORé J J,GARBOW B S,HILLSTROM K E.Testing unconstrained optimization software[J].ACM Transactions on Mathematical Software,1981,7(1):17-41.