Interactive implicit modelling

点击翻译按钮获取中文摘要

The research focus is on new methods for building computer models for application in computer assisted surgery, medical education, animation, and electronic games. A modelling methodology known as implicit modelling, describes a model as an iso-surface in a scalar field. This is similar to the idea of a set of contours in a height field. The collection of contours describes a surface. Solid models can be produced that exhibit several advantages over the more usual method that use triangle meshes. Implicit models are solid, it is easy to tell inside from outside and find for example contact between models of a human organ and a virtual knife. These models are very useful for describing both manufactured objects that have sharp edges and more organic objects. Using results of our earlier research it is possible to describe very complex models with very little data, enabling transmission over a network, thus making such models useful in cooperative situations where users are geographically far apart. For example virtual surgery. Earlier work allowed rapid prototyping of complex (topologically and geometrically) models from user sketches and we are still finding better methods of producing highly complex models from simple input. There are currently several disadvantages that prevent wide acceptance of implicit modelling. it is difficult to generate implicit models from scan data, and visualization is slow compared to mesh based methods. Research is required to explore new field functions that allow for better control over the shape of the model particularly during animation. In this work the intention is to investigate methods to overcome these problems to make implicit modelling a useful method in the application areas indicated.