In April 2016 Manchester eScholar was replaced by the University of Manchester’s new Research Information Management System, Pure. In the autumn the University’s research outputs will be available to search and browse via a new Research Portal. Until then the University’s full publication record can be accessed via a temporary portal and the old eScholar content is available to search and browse via this archive.

Related resources

Search for item elsewhere

University researcher(s)

Academic department(s)

On Quantitative Validation and Optimisation of Physically Based Deformable Models in Computer Graphics

Banks, Matthew

[Thesis]. Manchester, UK: The University of Manchester; 2020.

Access to files

FULL-TEXT.PDF (pdf)

Abstract

Physically based deformable models (PBDMs) in computer graphics (CG) simulate how virtual solid bodies respond under external loads in computer applications that must be computationally efficient enough so as to be user-interactive. Balance laws from the continuum mechanics theory govern the behaviour of solids. The modelling and simulation of this behaviour is traditionally performed using the finite element method (FEM), which solves to full accuracy and is therefore computationally expensive. This typically makes it unsuitable for IVEs and so different PBDMs are researched that are less expensive. These PBDMs make modelling assumptions that simplify the theory and their numerical results are less accurate than FEM. In this thesis we present a novel software framework that allows us to quantify the accuracy of any deformation history of any PBDM as a result of its modelling simplifications. In the majority of previous studies, validation is qualitative (through visual plausibility), which is necessarily user subjective. We develop a novel, objective, quantitative validation of deformation histories procedure (QVDH) to quantify the accuracy of any PBDM with an error between 0 and 1. We test QVDH in 3D cantilever and cloth scenarios, both of which are popular scenarios in the CG literature. The framework is shown to yield a high accuracy score for a simplified model that can be analytically derived from the reference model, indicating that the framework is reliable. We then extend the framework to optimise the model properties of PBDMs (that determine the material response) by minimising the error measured by QVDH. Results are in good agreement with analytically derived results, showing the effectiveness of the procedure. Finally, we use the framework to explore adaptive PBDMs - in particular PBDMs that switch to other PBDMs at runtime - and demonstrate how switching can successfully be implemented to increase the accuracy of a deformation according to QVDH. The software framework is sufficiently general to be applicable to a wide variety of deformation scenarios. It has replaceable components that can help to improve the quality of the QVDH procedure. We believe that QVDH has many uses beyond those explored in this thesis.