基于光流反馈的单目视觉三维重建

李秀智; 杨爱林; 秦宝岭; 贾松敏; 邱欢

doi:doi:10.3788/aos201535.0515001

光学学报, 2015, 35 (5): 0515001, 网络出版: 2015-04-28

基于光流反馈的单目视觉三维重建下载： 730次

Monocular Camera Three Dimensional Reconstruction Based on Optical Flow Feedback

论文大纲

李秀智 ^*杨爱林秦宝岭贾松敏邱欢

作者单位

北京工业大学电子信息与控制工程学院, 北京 100124

机器视觉三维重建光流场景流统一计算设备架构 machine vision three dimensional reconstruction optical flow scene flow compute unified device architecture

AI 词云图 AI一句话精读 AI短摘要

注：本部分内容由 AI 自动生成，请您知悉。

摘要

提出一种基于光流反馈的单目视觉三维(3D)重建方法，实现对场景快速、准确的3D 立体化建模。由帧间光流场建立更为稳健的同名像点匹配关系，同时运用五点算法估计摄像机的相对位姿，以构建稀疏点云和初始网格。从运动视觉分析的角度寻求多视重构的求解方法，将重建模型反馈至重建过程，用各视图像的偏差驱动模型变形。将粗略、不准确的原始网格曲面经过致密的非刚性变形，调整至精确的曲面。在统一计算设备架构下，利用图形处理器对光流算法进行并行加速，显著提高了重构算法运行的实时性。室内真实场景下的重建结果证明了所提算法的可行性与准确性。

Abstract

A monocular three dimensional(3D) reconstruction technique based on optical flow feedback is proposed to achieve fast and accurate 3D stereoscopic modeling in the real scene. Corresponding pixel pairs are robustly matched by inter-frame optical flow fields and the five-point algorithm is employed to determine relative pose of the moving camera, therefore sparse point cloud is generated and initial crude mesh is built. In the proposed method, multi-view reconstruction is implemented from perspective of vision method on motion analysis. The reconstruction model is fed-back to the reconstruction process and the model is deformed by utilizing the bias-driven of each view. The coarse and inaccurate original mesh surface is adjusted to the exact surface through a dense non- rigid deformation. Under the compute unified device architecture, the optical flow algorithm is optimized in parallel mode by using the graphic processing unit hardware and real- time performance of the reconstruction algorithm is significantly improved. The experimental results obtained in realistic indoor scenario demonstrate the effectiveness and accuracy of the proposed algorithm.

参考文献

[1] 杨玉峰, 吴振森, 曹运华. 基于三维重建理论的目标光谱散射特性研究[J]. 光学学报, 2012, 32(9): 0929001.

Yang Yufeng, Wu Zhensen, Cao Yunhua. Research on the spectral scattering of target based on three-dimensional reconstruction theory[J]. Acta Optica Sinica, 2012, 32(9): 0929001.

[2] 贾松敏, 王可, 李秀智, 等. 基于变分模型的单目视觉三维重建方法[J]. 光学学报, 2014, 34(4): 0415002.

Jia Songmin, Wang Ke, Li Xiuzhi, et al.. Monocular camera three dimensional reconstruction based on variation model[J]. Acta Optica Sinica, 2014, 34(4): 0415002.

[3] X Li, S Jia, W Cui, et al.. Consistent map building by a mobile robot equipped with stereo sensor and LRF[C]. IEEE International Conference on Computer Science and Automation Engineering, 2011, 3: 100-104.

[4] T Brox, J Malik. Large displacement optical flow: descriptor matching in variational motion estimation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2011, 33(03): 500-513.

[5] S Vedula, S Baker, P Rander, et al.. Three-dimensional scene flow[C]. Proceedings of the International Conference on Computer Vision, Corfu, 1999: 722-729.

[6] S Vedula, P pomder, R collins, et al.. Three- dimensional scene flow[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2005, 27(3): 475-480.

[7] Y Zhang, C Kambhamettu. On 3D scene flow and structure recovery from multi-view image sequences[J]. IEEE Transactions on Systems, Man, and Cybernetics (Part B), 2003, 33(4): 592-606.

[8] R Li, S Sclaroff. Multi-scale 3D scene flow from binocular stereo sequences[J]. Computer Vision and Image Understanding, 2008, 110(1): 75-90.

[9] F Huguet, F Devernay. A variational method for scene flow estimation from stereo sequences[C]. Proceedings of the International Conference on Computer Vision, Los Alamitos, 2007: 1-7.

[10] Richard A. Newcombe, A J Davision. Live dense reconstruction with a single moving camera[C]. Computer Vision and Pattern Recongnition (CVPR), 2010: 1498-1505.

[11] X Li, S Jia, K Wang, et al.. Scene flow-based environment 3D digitalization for mobile robot navigation[C]. Advanced Robotics, 2012, 26(3): 1521-1536.

[12] B Horn, B Schunck. Determining optical flow[C]. International Joint Conference on Artificial Intelligence, 1981: 319-331.

[13] 李秀智, 尹晓琳, 贾松敏, 等. 改进的TV-L1平滑光流估计[J]. 光学学报, 2013, 33(10): 1015002.

Li Xiuzhi, Yin Xiaolin, Jia Songmin, et al.. Improved TV-L1 algorithm for smooth optical flow[J]. Acta Optica Sinica, 2013, 33(10): 1015002.

[14] C Zach, T Pock, H Bischof. A Duality Based Approach for Real-Time TV-L1 Optical Flow[M]. Berlin: Springer-Verlag, 2007: 214-223.

[15] T Pock, M Urschler, C Zack, et al.. A duality based algorithm for TV- L1- optical- flow image registration[C]. International Conference on Medical Image Computing and Computer Assisted Intervention, 2007: 511-518.

[16] D Nister. An efficient solution to the five- point relative pose problem[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2004, 26(6): 756-770.

[17] N Snavely, S M Seitz, R Szeliski. Photo tourism: exploring image collections in 3D[C]. ACM Transactions on Graphics, 2006, 25(3): 835-846.

[18] N Snavely, S M Seitz, R Szeliski. Modeling the world from internet photo collections[J]. International Journal of Computer Vision, 2008, 80(2): 189-210.

[19] I Tobor, P Reuter, C Schlick. Efficient reconstruction of large scattered geometric datasets using the partition of unity and radial basis functions[J]. WSCG, 2004, 12(1-3): 467-474.

[20] J Bloomenthal. An Implic1it Surface Polygonizer, Graphics Gems IV[M]. San Diego: Academic Press, 1994: 324-349.

[21] 张峰, 史利民, 孙凤梅, 等. 一种基于图像的室内场景自动三维重建系统[J]. 自动化学报, 2010, 36(5): 625-633.

Zhang Feng, Shi Limin, Sun Fengmei, et al.. An Image Based 3D reconstruction system for large indoor scenes[J]. Acta Automation Sinica, 2010, 36(5): 625-633.

李秀智, 杨爱林, 秦宝岭, 贾松敏, 邱欢. 基于光流反馈的单目视觉三维重建[J]. 光学学报, 2015, 35(5): 0515001. Li Xiuzhi, Yang Ailin, Qin Baoling, Jia Songmin, Qiu Huan. Monocular Camera Three Dimensional Reconstruction Based on Optical Flow Feedback[J]. Acta Optica Sinica, 2015, 35(5): 0515001.

基于光流反馈的单目视觉三维重建下载： 730次

关于本站 Cookie 的使用提示

全站搜索

基于光流反馈的单目视觉三维重建 下载： 730次

相关论文

相关资讯

关于本站 Cookie 的使用提示

全站搜索

基于光流反馈的单目视觉三维重建下载： 730次