A secondary parametric model for CSG.

Departamento de Sistemas Informáticos y Computación, Universidad Politécnica de Valencia, Camino de Vera, s/n, 46022 Valencia, Spain
Computers & Graphics 01/1992; 16:369-373. DOI: 10.1016/0097-8493(92)90023-O
Source: DBLP

ABSTRACT In this paper, a new method based on a quadtree decomposition is proposed for rendering CSG modeled solids. A surface parametric description is stored in a quadtree (VST) whose nodes represent parametric patches. Object description allows a non-polygonal boundary representation that provides more accurate point sampling information for the rendering process. The visible surface tree (VST) is built and updated at CSG edition time. Traversal is performed when a realistic object image is required.

  • [Show abstract] [Hide abstract]
    ABSTRACT: Computer graphics have long been used to help people visualise complex entities such as machine components and assemblies. Systems such as Computer-Aided Design (CAD) applications allow the creation of models and designs for objects by providing interactive tools that can be manipulated by designers. An important facet of interactive design is that it should be possible to see the effects of changes made on models immediately, i.e. in real-time. In this paper we are interested in applications where renderings of complex objects, modelled using Constructive Solid Geometry (CSG), can be updated quickly enough to allow interactive use. The algorithms developed in this paper to allow real-time rendering of CSG-based objects require no specialised hardware or software to perform their tasks. They merely need a display adapter with stencil buffer support, and a graphics language capable of manipulating the colour, depth and stencil buffers provided by the display hardware. They can be implemented on low-cost PC- level systems and still prove capable of generating the required number of frames per second to allow real-time screen updates of CSG-modelled objects. Through empirical studies and observation we have identified that the algorithms developed here provide better performance than equivalent algorithms.