TY - GEN
T1 - Optimizing on real-time fluid 3D effect in mobile environment
AU - Lu, Jue Min
AU - Piao, Jin Chun
AU - Kim, Shin Dug
PY - 2015/1/8
Y1 - 2015/1/8
N2 - Simulating fluids on screen has been getting many spotlights because it is one of the most challenging tasks in computer science area. So we developed a real-time fluid simulation on android platform with SPH fluid algorithm and Bullet physics engine, and then rendered it with OpenGL|ES. Because of the limitation of smartphone environment, the simulation does not show a good performance. We suggest three methods to improve the performance of the demo. Firstly, we enable parallel computing with ARM NEON to accelerate mathematical computation. Secondly, we simplify the SPH algorithm with particle-forming technique. Finally, we optimize rendering skill with SSFR technique. We carry out experiments to validate the methods we suggest. The experiments results show that the first method improved the performance by about 26% to 40%, the second method improved the performance by about 50%, and the third method helped us represent a more smooth and natural fluid surface. Therefore, these methods are proved to be able to improve the performance of the real-time 3D fluid effect simulation significantly.
AB - Simulating fluids on screen has been getting many spotlights because it is one of the most challenging tasks in computer science area. So we developed a real-time fluid simulation on android platform with SPH fluid algorithm and Bullet physics engine, and then rendered it with OpenGL|ES. Because of the limitation of smartphone environment, the simulation does not show a good performance. We suggest three methods to improve the performance of the demo. Firstly, we enable parallel computing with ARM NEON to accelerate mathematical computation. Secondly, we simplify the SPH algorithm with particle-forming technique. Finally, we optimize rendering skill with SSFR technique. We carry out experiments to validate the methods we suggest. The experiments results show that the first method improved the performance by about 26% to 40%, the second method improved the performance by about 50%, and the third method helped us represent a more smooth and natural fluid surface. Therefore, these methods are proved to be able to improve the performance of the real-time 3D fluid effect simulation significantly.
UR - http://www.scopus.com/inward/record.url?scp=84926155312&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84926155312&partnerID=8YFLogxK
U2 - 10.1145/2701126.2701147
DO - 10.1145/2701126.2701147
M3 - Conference contribution
AN - SCOPUS:84926155312
T3 - ACM IMCOM 2015 - Proceedings
BT - ACM IMCOM 2015 - Proceedings
PB - Association for Computing Machinery, Inc
T2 - 9th International Conference on Ubiquitous Information Management and Communication, ACM IMCOM 2015
Y2 - 8 January 2015 through 10 January 2015
ER -