Abstract: To speed up the reconstruction of 3D dynamic scenes in an ordinary hardware platform, we propose an efficient framework to reconstruct 3D dynamic objects using a multiscale-contour-based interpolation from multi-view videos. Our framework takes full advantage of spatio-temporal-contour consistency. It exploits the property to interpolate single contours, two neighboring contours which belong to the same model, and two contours which belong to the same view at different times, corresponding to point-, contour-, and model-level interpolations, respectively. The framework formulates the interpolation of two models as point cloud transport rather than non-rigid surface deformation. Our framework speeds up the reconstruction of a dynamic scene while improving the accuracy of point-pairing which is used to perform the interpolation. We obtain a higher frame rate, spatio-temporal-coherence, and a quasi-dense point cloud sequence with color information. Experiments with real data were conducted to test the efficiency of the framework.

基于多尺度轮廓插值生成准密集时变点云模型序列

[1]Ahmed, N., Junejo, I.N., 2013. A system for 3D video acquisition and spatio-temporally coherent 3D animation reconstruction using multiple RGB-D cameras. Int. J. Signal Process. Image Process. Patt. Recogn., 6(2):113-128.

[2]Allain, B., Franco, J.S., Boye, R.E., 2015. An efficient volumetric framework for shape tracking. IEEE Conf. on Computer Vision and Pattern Recognition, p.268-276.

[3]Arita, D., Taniguchi, R., 2001. RPV-II: a stream-based real-time parallel vision system and its application to real-time volume reconstruction. 2nd Int. Workshop on Computer Vision Systems, p.174-189.

[4]Baumgart, B.G., 1974. Geometric Modeling for Computer Vision. PhD Thesis, Stanford University, Stanford, USA.

[5]Bilir, S.C., Yemez, Y., 2012. Non-rigid 3D shape tracking from multiview video. Comput. Vis. Image Understand., 116(11):1121-1134.

[6]Borovikov, E., Sussman, A., Davis, L., 2003. A high performance multi-perspective vision studio. 17th Annual Int. Conf. on Supercomputing, p.348-357.

[7]Cheung, G.K.M., Kanade, T., Bouguet, J.Y., 2000. A real time system for robust 3D voxel reconstruction of human motions. IEEE Conf. on Computer Vision and Pattern Recognition, p.714-720.

[8]Cheung, G.K.M., Baker, S., Kanade, T., 2003. Visual hull alignment and refinement across time: a 3D reconstruction algorithm combining shape-from-silhouette with stereo. IEEE Computer Society Conf. on Computer Vision and Pattern Recognition, p.375-382.

[9]Díaz-Más, L., Muñoz-Salinas, R., Madrid-Cuevas, F.J., et al., 2010. Shape from silhouette using Dempster-Shafer theory. Patt. Recog., 43(6):2119-2131.

[10]Duckworth, T., Roberts, D.J., 2011. Accelerated polyhedral visual hulls using OpenCL. IEEE Virtual Reality Conf., p.203-204.

[11]Franco, J.S., Boyer, E., 2009. Efficient polyhedral modeling from silhouettes. IEEE Trans. Patt. Anal. Mach. Intell., 31(3):414-427.

[12]Furukawa, Y., Ponce, J., 2009. Carved visual hulls for image-based modeling. Int. J. Comput. Vis., 81(1):53-67.

[13]Furukawa, Y., Ponce, J., 2010. Accurate, dense, and robust multiview stereopsis. IEEE Trans. Patt. Anal. Mach. Intell., 32(8):1362-1376.

[14]Haro, G., 2012. Shape from silhouette consensus. Patt. Recogn., 45(9):3231-3244.

[15]Hasler, N., Rosenhahn, B., Thormahlen, T., et al., 2009. Markerless motion capture with unsynchronized moving cameras. IEEE Conf. on Computer Vision and Pattern Recognition, p.224-231.

[16]Hauswiesner, S., Khlebnikov, R., Steinberger, M., et al., 2012. Multi-GPU image-based visual hull rendering. 12th Eurographics Symp. on Parallel Graphics and Visualization, p.119-128.

[17]Hofmann, M.H., Davrila, M.G., 2009. Multi-view 3D human pose estimation combining single-frame recovery, temporal integration and model adaptation. IEEE Conf. on Computer Vision and Pattern Recognition, p.2214-2221.

[18]Huang, C.H., Lu, D.M., Diao, C.Y., 2013. Accelerated visual hulls of complex objects using contribution weights. Proc. 7th Int. Conf. on Image and Graphics, p.685-689.

[19]Huang, C.H., Lu, D.M., Diao, C.Y., 2014a. A point cloud representation using plane-space-local-area-color- consistency. J. Comput.-Aided Des. Comput. Graph., 26(8):1297-1303 (in Chinese).

[20]Huang, C.H., Boyer, E., Navab, N., et al., 2014b. Human shape and pose tracking using keyframes. IEEE Conf. on Computer Vision and Pattern Recognition, p.3446-3453.

[21]Kanaujia, A., Haering, N., Taylor, G., et al., 2011. 3D human pose and shape estimation from multi-view imagery. IEEE Computer Vision and Pattern Recognition Workshops, p.49-56.

[22]Kim, D., Dahyot, R., 2012. Bayesian shape from silhouettes. Int. Workshop on Multimedia Understanding Through Semantics, Computation, and Learning, p.78-89.

[23]Laurentini, A., 1994. The visual hull concept for silhouette-based image understanding. IEEE Trans. Patt. Anal. Mach. Intell., 16(2):150-162.

[24]Li, K., Dai, Q.H., Xu, W.L., 2011. Markerless shape and motion capture from multiview video sequences. IEEE Trans. Circ. Syst. Video Technol., 21(3):320-334.

[25]Liu, Y.B., Dai, Q.H., Xu, W.L., 2010. A point-cloud-based multiview stereo algorithm for free-view-point video. IEEE Trans. Vis. Comput. Graph., 16(3):407-418.

[26]Matsuyama, T., Wu, X.J., Takai, T., et al., 2004. Real-time dynamic 3-D object shape reconstruction and high-fidelity texture mapping for 3-D video. IEEE Trans. Circ. Syst. Video Technol., 14(3):357-369.

[27]Matusik, W., Buehler, C., Raskar, R., et al., 2000. Image-based visual hulls. ACM Special Interest Group on Computer Graphics, p.369-374.

[28]Nakajima, H., Makihara, Y., Hsu, H., et al., 2012. Point cloud transport. 21st Int. Conf. on Pattern Recognition, p.3803-3806.

[29]Nakazawa, M., Mitsugami, I., Makihara, Y., et al., 2012. Dynamic scene reconstruction using asynchronous multiple Kinects. 21st Int. Conf. on Pattern Recognition, p.469-472.

[30]Perez, J.M., Aledo, P.G., Sanchez, P.P., 2012. Real-time voxel-based visual hull reconstruction. Microprocess. Microsyst., 36(5):439-447.

[31]Raeesi N., M.R., Wu, Q.M.J., 2010. A complete visual hull representation using bounding edges. 11th Pacific-Rim Conf. on Multimedia, p.171-182.

[32]Taneja, A., Ballan, L., Pollefeys, M., 2011. Modeling dynamic scenes recorded with freely moving cameras. 10th Asian Conf. on Computer Vision, p.613-626.

[33]Vlasic, D., Peers, P., Baran, I., et al., 2009. Dynamic shape capture using multi-view photometric stereo. ACM Trans. Graph., 28(5):174.

[34]Wang, S.Y., Yu, H.M., 2012. Convex relaxation for a 3D spatiotemporal segmentation model using the primal-dual method. J. Zhejiang Univ.-Sci. C (Comput. & Electron), 13(6):428-439.

[35]Wu, X.J., Takizawa, O., Matsuyama, T., 2006. Parallel pipeline volume intersection for real-time 3D shape reconstruction on a PC cluster. IEEE Int. Conf. on Computer Vision Systems, p.1-4.

[36]Xia, D., Li, D.H., Li, Q.G., 2011. A novel approach for computing exact visual hull from silhouettes. Optik, 122(24):2220-2226.

[37]Zhang, Z., Seah, H.S., Quah, C.K., et al., 2011. A multiple camera system with real-time volume reconstruction for articulated skeleton pose tracking. 17th Int. Multimedia Modeling Conf., p.182-192.