Supervisor: Alan Dorin
Creation of realistic images using computer algorithms is a constant challenge. In particular the generation of sinewy networks, which grow based on a range of factors in nature, is very difficult to reproduce correctly by way of computer graphics. Factors such as surrounding objects and energy from light sources are implicit in the growth of real sinewy networks such as plant life, however they are hard to simulate in software. Previously Greene [7] created a method that generated a virtual sinewy network by simulating light and surrounding objects. However, his method had certain limitations, the largest of which was the inability to model more than one source of light. While this limitation did not lessen the realism of his images, it was not possible to create more complicated results using his method. This project aims to develop a Sinewy Network Generator (SiNGe), a piece of software that has the ability to generate sinewy networks while simulating light and surrounding objects with great variability. In particular the software will support multiple light sources, and will be able handle a wider range of objects for the sinewy network to interact with as it is being generated. It will also be possible to change the priorities of the growing sinewy network to better reflect different environments, and to produce data that will allow the sinewy network to either by rendered as a single image or as an animation.
Code |
DocumentsThesis.pdfFinalPresentation.ppt Literature_Review.pdf Research_Proposal.pdf Latex Source Latex Image Files (15 MB) |
Image Log |