Frank C Langbein
Ex Tenebris Scientia
Updates: Research

I am a member of the Geometric Computing and Computer Vision research group at the School of Computer Science, Cardiff University. I received a diploma in mathematics from Stuttgart University in 1998 and a PhD from Cardiff University in 2003. My research interests lie in the areas of geometric, solid and physical modelling, computational geometry, differential geometry and related topics in computer graphics. I have been working on reverse engineering geometric models with a focus on beautification and design intent detection, approximate symmetries and pattern recognition, geometric constraints, mesh processing, curves and surfaces, and point-based modelling. I am interested in applying this in engineering and science, especially in mechanical engineering, quantum and nano technology, and life sciences. I am a member of the American Mathematical Society, the IEEE, the Geometric Modelling Society and the Research Institute for Visual Computing where I am co-leading the sub-programme on vision-based geometric modelling and the interface with science.

Academia.EDU | Academici | ResearchGATE

Current Focus

Currently I am mainly working on low-level representations of shape for modelling and simulation operations, but also describing shape in terms of high level properties for creating, editing and analysing geometric models. This is aimed at computer-aided design applications and design intent for modelling conventional devices, but also for modelling and controlling quantum systems to build novel quantum devices.

Current projects include:

Some more specific themes are described below and for more details see the sections to the left, sorted roughly by the type of objects considered.

Understanding Shape

Understanding, describing and handling the shape and specific geometric and functional properties of physical objects and building suitable computational models for this purpose has the potential to simplify processing and representing shape and may improve human interaction with such models. I have been working on reverse engineering geometric models with the particular aim of automatically generating high level representations of the shapes, which exhibit their intended geometric properties. The models have either been reconstructed by measuring real objects or are simply basic CAD models without any explicit information about their design or intended properties. This work resulted in a system for detecting a likely design intent description of approximate geometric models and beautification of such models based on various algorithms for detecting approximate symmetries and other regularities, selecting suitable intended regularities, and enforcing such regularities on a model using geometric constraints. This system may be used to improve, edit and analyse approximate geometric models in terms of higher level geometric properties. The ultimate goal of this work is to devise a system which can automatically produce different interpretations of a model (in terms of suitable notions of regularity relevant to the desired interpretation or understanding) in real-time to enable an engineer to easily create, modify and analyse such models without restrictions based on how it has been created initially or how it is currently interpreted due to a specific representation or list of properties used to describe the shape. This could be done not only for solid physical objects, but for any geometric model of a physical system to help understand their operation and create and modify them for a particular purpose.

Related projects include:

Modelling the Physical World

I am also interested in modelling and simulating physical systems using discrete models suitable for computation. An interesting approach to computing with geometric objects is to represent them as discrete point sets where each point is a sample of the underlying geometry. Often this is referred to as meshless or point-based processing, where the use of any connectivity information has been minimised. This computationally very simple representation has been used for rendering, other modelling operations and simulations. In general low-level representations may allow efficient processing. However, additional high level descriptions of the point set's properties may make it simpler to edit the point set and at the same time allow us to use the point representation for simluations and other analysis. Such descriptions may at least be partially computed automatically from the low-level representation. Work on this resulted in an approach to sampling manifolds and there is ongoing work on using point sets for simulations. This is aimed particularly at modelling quantum systems for applications in nanotechnology.

Related projects include:

As time wore along, his absorption in the irregular wall and ceiling of his room increased; for he began to read into the odd angles a mathematical significance which seemed to offer vague clues regarding their purpose.
H.P. Lovecraft, The Dreams in the Witch House

Cite as Research, [21/July/2009, 10:17].
[Contact Details]