“3D reconstruction of detail change on dynamic non-rigid objects” by Taneda, Suda, Maejima and Morishima

  • ©Daichi Taneda, Hirofumi Suda, Akinobu Maejima, and Shigeo Morishima

  • ©Daichi Taneda, Hirofumi Suda, Akinobu Maejima, and Shigeo Morishima

Conference:


Type(s):


Title:

    3D reconstruction of detail change on dynamic non-rigid objects

Presenter(s)/Author(s):



Abstract:


    3D reconstruction of a moving object is notable technique in fields such as digital archives or entertainments. [de Aguiar et al. 2008] is able to reconstruct motion and spatio-temporally coherent time-varying geometry using by eight multi-view video and a range scanner. However, they couldn’t capture the numerous high-frequency folds of garments, it make the realism of the re-constructed dynamic models reduce. On the other hand, [Hernández et al. 2007] proposed the technique to reconstruct detailed 3D shape of moving cloth including high frequency folds by using one camera. They also use three primary colored lights to acquire the shading information independently at the same time. This property made it possible to estimate surface normals on the object from the video sequence by using photometric stereo. However, there is a defect that the area which can be reconstructed is only limited to frontal surface.

References:


    1. de Aguiar, E., Stoll, C., Theobalt, C., Ahmed, N., Seidel, H.-P., and Thrun, S. 2008. Performance capture from sparse multi-view video. ACM Transactions on Graphics 27, 98.
    2. Hernandez, C., Vogiatzis, G., Brostow, G. J., Stenger, B., and Cipolla, R. 2007. Non-rigid photometric stereo with colored lights. In Proc. of ICCV.


Additional Images:

©Daichi Taneda, Hirofumi Suda, Akinobu Maejima, and Shigeo Morishima ©Daichi Taneda, Hirofumi Suda, Akinobu Maejima, and Shigeo Morishima

ACM Digital Library Publication:



Overview Page: