Fast, realistic and controllable smoke simulation

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Highly detailed smoke simulation in large scale environment is extremely time-consuming. In this paper, we describe a very fast simulator which can produce realistic and controllable smoke animations. Our simulator is entirely implemented on GPU. The most expensive step in traditional method involves a Poisson equation. We develop a very efficient Poisson solver that can be perfectly implemented on GPU. This solver does not need any global matrix and moving obstacles in the volume can be easily handled without extra time for building matrix. For users to design artistic effects, we develop a controlling method with line sketches. The line sketches with attached properties are translated by our simulator to a direction field and a velocity field. These two fields can help produce various artistic effects. Experiments have shown our method can generate realistic smoke animations in various situations.

Original languageEnglish
Title of host publicationProceedings - 2015 International Conference on Virtual Reality and Visualization, ICVRV 2015
EditorsZhong Zhou, Weiliang Meng, Junfeng Yao, Xiaopeng Zhang, Xun Luo
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages197-200
Number of pages4
ISBN (Electronic)9781467376730
DOIs
StatePublished - 9 May 2016
Event5th International Conference on Virtual Reality and Visualization, ICVRV 2015 - Xiamen, Fujian, China
Duration: 17 Oct 201518 Oct 2015

Publication series

NameProceedings - 2015 International Conference on Virtual Reality and Visualization, ICVRV 2015

Conference

Conference5th International Conference on Virtual Reality and Visualization, ICVRV 2015
Country/TerritoryChina
CityXiamen, Fujian
Period17/10/1518/10/15

Keywords

  • Control simulation
  • GPU
  • Poisson solver
  • Sketch
  • Smoke simulation

Fingerprint

Dive into the research topics of 'Fast, realistic and controllable smoke simulation'. Together they form a unique fingerprint.

Cite this