Mohammad Rouhani: Surface Reconstruction and Registration using Implicit Functions

Implicit functions are among the most convenient tools for surface description as they do not require any parameterization on the given point cloud. In this talk, I will present a fast and flexible reconstruction technique based on implicit b-splines. As the basis functions are locally supported, the optimal surface parameters can be found by solving a sparse system of equations. This method can be accelerated by reducing the dimension, while the outcome can be further controlled by user interaction and regularization. Moreover, a novel weighting technique has been introduced in order to blend small patches and smooth them just in the overlapping regions. The whole framework is very fast and efficient and can handle large cloud of points with low computational cost. Finally, I will show how this representation can be exploited for solving non-rigid surface registration in motion capture.

Mohammad Rouhani received the master degrees in applied mathematics from Sharif University of Technology, Tehran, Iran, in 2006. Having experienced two years of lecturing in Computer Science, he joined the Computer Vision Center in Barcelona, Spain, where he received the Ph.D. degree in Computer Science, in 2012. After holding a research position in ICV Lab at Imperial College London, U.K., he joined Morpheo team at INRIA Grenoble. He is currently a postdoc researcher at Titane team in INRIA. His research interests include 3D Computer Vision and Computer Graphics, including shape modeling, surface reconstruction, deformation modeling, object regis- trations well as machine learning techniques for 3D pose estimation.