On Prism-based Motion Blur and Locking-proof Tetrahedra
Licentiatavhandling, 2021

Motion blur is an important visual effect in computer graphics for both real-time, interactive, and offline applications. Current methods offer either slow and accurate solutions for offline ray tracing applications, or fast and inaccurate solutions for real-time applications.

This thesis is a collection of three papers, two of which address the need for motion blur solutions that cater to applications that need to be accurate and as well as interactive, and a third that addresses the problem of locking in standard FEM simulations. In short, this thesis deals with the problem of representing continuous motion in a discrete setting.

In Paper I, we implement a GPU based fast analytical motion blur renderer. Using ray/triangular prism intersections to determine triangle visibility and shading, we achieve interactive frame rates.

In Paper II, we show and address the limitations of using prisms as approximations of the triangle swept volume. A hybrid method of prism intersections and time-dependent edge equations is used to overcome the limitations of Paper I.

In Paper III, we provide a solution that alleviates volumetric locking in standard Neo-Hookean FEM simulations without resorting to higher order interpolation.

continuous collision detection

locking

GPU

parallel

real time

FEM

motion blur

prism

Online
Opponent: Marco Tarini, University of Milan, Italy

Författare

Mads Rönnow

Chalmers, Data- och informationsteknik, Interaktionsdesign (Chalmers)

Rønnow, M. J. L. Assarsson, U. Fratarcangeli, M. Fast Analytical Motion Blur with Transparency

Rønnow, M. J. L. Assarsson, U. Fratarcangeli, M. Improved Accuracy for Prism-based Motion Blur

Frâncu, M. Asgeirsson, A. Rønnow, M. J. L. Erleben, K. Locking-Proof Tetrahedra

Interaktiva 3D deformerbara kroppars design

Vetenskapsrådet (VR), 2016-01-01 -- 2018-12-31.

Fotorealistisk rendering för realtid

Vetenskapsrådet (VR), 2014-01-01 -- 2021-12-31.

Styrkeområden

Informations- och kommunikationsteknik

Ämneskategorier

Datavetenskap (datalogi)

Utgivare

Chalmers tekniska högskola

Online

Online

Opponent: Marco Tarini, University of Milan, Italy

Mer information

Senast uppdaterat

2021-03-10