PARALLEL PROCESSING IMAGE SYNTHESIS (COMPUTER GRAPHICS, ALGORITHMS)
Abstract
For many current applications of computer graphics, realistic rendition and speed of generation are two important goals. Realistic pictures can be generated by increasing the resolution of the image array and by using more complicated visible-surface algorithms and shading models. This in turns will increase the computational load dramatically. In order to meet this challenge, parallel computers and parallel algorithms must be employed to speedup the computation. In this research, we study the parallel architectures and algorithms for removing hidden surfaces from three dimensional scenes represented by the constructive solid geometry techniques. Our focus is on the parallel ray casting algorithms. The architectures which are suitable to execute the parallel algorithms are discussed and analyzed. We identify the key factors which could degrade the performance of the parallel algorithm and propose methods to overcome these factors. We demonstrate that good performance data can be achieved by the parallel algorithms and architectures proposed in this thesis.
Degree
Ph.D.
Subject Area
Computer science
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.