Return to search

Spherical Terrains Compared : A Standardized Performance Comparison Between Heightmaps and Marching Cubes

Background. Games are becoming increasingly large, and that includes space games. They require new features to be developed and planets to be generated. Heightmap- and marching cubes-algorithms, allow for the generation of spherical objects and planets when adapted. With the help of noise and these algorithms,spherical terrains are created in this project the performance of the individual solutions is measured. Objectives. It is the scope of this project to implement two planet generators, one based on heightmap technology and another based on the marching cube algorithm.These are then compared in terms of performance to see the differences in memory consumption and which one is faster.  Methods. The two techniques, planetary marching cubes, and a planetary heightmap-based solution are implemented in the Unity Engine. These generators then generate a set of different sized terrains multiple times and then average the results to gather performance data. This data is then used to compare the two solutions. This yields a comparable result between the two techniques which can then be analysed.  Results. The resulting measurements show the Marching cubes implementation uses a higher consumption of both VRAM and RAM than the heightmap solution, which is very memory efficient. Furthermore, the generation time of the heightmap solution is significantly faster than the marching cubes implementation. Conclusions. It is concluded that the planetary marching cubes algorithm utilizes more resources, consumes time to generate and is slower than the heightmap-based solution. The planetary marching cubes algorithm does however support additional features.

Identiferoai:union.ndltd.org:UPSALLA1/oai:DiVA.org:bth-21757
Date January 2021
CreatorsLeo, Christian, Hansen, Oskar
PublisherBlekinge Tekniska Högskola, Institutionen för datavetenskap
Source SetsDiVA Archive at Upsalla University
LanguageEnglish
Detected LanguageEnglish
TypeStudent thesis, info:eu-repo/semantics/bachelorThesis, text
Formatapplication/pdf
Rightsinfo:eu-repo/semantics/openAccess

Page generated in 0.0018 seconds