Self-similarities in Urban Data Reconstruction

--------------------------------------------------------------------
                ***Joint Seminar***
--------------------------------------------------------------------
The Hong Kong University of Science & Technology

Department of Computer Science and Engineering
Center for Visual Computing and Image Science
--------------------------------------------------------------------

Speaker:	Prof. Andrei SHARF
		SIAT Institute, Shenzhen, China
		Ben-Gurion University, Israel

Title:		"Self-similarities in Urban Data Reconstruction"

Date: 		Monday, 12 April 2010

Time: 		10:30am - 11:30am

Venue: 		Room 3401 (via lifts 17/18)
		The Hong Kong University of Science & Technology


Abstract:

Today's scanning technologies allow fast scanning of urban scenes. Such
rapid acquisition incurs imperfections: large regions remain missing,
significant variation in sampling density is common, and the data is often
corrupted with noise and outliers.

Buildings often exhibit large scale repetitions and self-similarities. To
consolidate the imperfect data, our key observation is that the same
geometry, when scanned multiple times over reoccurrences of instances, allow
application of a simple and effective non-local filter. The multiplicity of
the geometry is fused together and projected into a base-geometry that is
denoised to consolidate reoccurrences.

Next, we introduce an interactive tool which enables a user to quickly
assemble an architectural model directly over a 3D scanned urban scene. The
user loosely defines and manipulates simple building blocks, which we call
SmartBoxes that snap to their proper locations. The building blocks are
smart in the sense that their locations and sizes are automatically adjusted
on-the-fly to fit to the point data, while respecting contextual relations
with nearby similar blocks.


******************
Biography:

ANDREI SHARF is currently a visiting professor at SIAT Institute, Shenzhen,
China 2009-2010. He will join the Computer Science department at Ben-Gurion
University Israel, starting in October.

He received a B.Sc. in Computer Science from the Technion Institute in 1999,
and an M.Sc in Computer Science from the Tel-Aviv University in 2002. During
this time he spent four years working at a start-up software company.  His
PhD on "Surface Reconstruction Techniques for Imperfect Raw Data" in 2007
was carried out under the supervision of Daniel Cohen-Or and Ariel Shamir.
Sharf did his Post-Doc with Prof. Nina Amenta at UC-Davis in 2007-2009,
exploring topics in shape registration, computational geometry and parallel
data structures on the GPU.

His research interests include surface reconstruction, geometric modeling,
interactive applications, shape analysis and parallel algorithms. He is
currently working on surface deformation techniques, shape denoising and GPU
related data structures.