Fire Sprite Animation Using Fire-Flake Texture and Artificial Motion Blur

In this paper, we propose a sprite animation synthesis technique that can efficiently represent the fire-flake effects seen in the natural phenomenon of fire. The proposed method uses the actual fire video or animated fire video as inputs and performs the following steps: 1) Extraction of feature vectors that can predict the direction of flame from image, 2) calculation of artificial buoyancy field, 3) creation and advection of fire-flake texture, 4) calculation of artificial motion blur using buoyancy flow, and 5) high quality composition. First, we detect the edges from the image and calculate the feature vectors needed to calculate the artificial buoyancy field. The computed 2D feature vectors are integrated into the Navier-Stokes equation and used to calculate the buoyancy field, which generates and advects anisotropic fire-flake textures. Finally, we apply artificial motion blur according to buoyancy direction to improve composition result of sprite animation. As a result, this method is based on image synthesis, which is faster than the existing 3D simulation-based approach. Experimental results show that high quality results can be easily and reliably obtained. In addition, since the final result is a sprite animation format, it can be easily used in existing game engines.

[1]  P. Burt.  Fast filter transform for image processing , 1981 .

[2]  N. Thürey,et al.  Data-driven synthesis of smoke flows with CNN-based feature descriptors , 2017, ACM Trans. Graph..

[3]  Matthias Müller,et al.  Position based dynamics , 2007, J. Vis. Commun. Image Represent..

[4]  Frank Losasso,et al.  A fast and accurate semi-Lagrangian particle level set method , 2005 .

[5]  Markus H. Gross,et al.  Particle-based fluid simulation for interactive applications , 2003, SCA '03.

[6]  Christopher Horvath,et al.  Directable, high-resolution simulation of fire on the GPU , 2009, SIGGRAPH '09.

[7]  Bernd Hamann,et al.  Real-time procedural volumetric fire , 2007, SI3D.

[8]  Ronald Fedkiw,et al.  Wrinkled flames and cellular patterns , 2007, SIGGRAPH 2007.

[9]  Randima Fernando,et al.  GPU Gems: Programming Techniques, Tips and Tricks for Real-Time Graphics , 2004 .

[10]  Ken Perlin,et al.  Accelerating Eulerian Fluid Simulation With Convolutional Networks , 2016, ICML.

[11]  Renato Pajarola,et al.  Predictive-corrective incompressible SPH , 2009, ACM Trans. Graph..

[12]  Chang-Hun Kim,et al.  Visual simulation of fire-flakes synchronized with flame , 2017, The Visual Computer.

[13]  Jos Stam,et al.  Stable fluids , 1999, SIGGRAPH.

[14]  Chang-Hun Kim,et al.  Physics-inspired controllable flame animation , 2016, The Visual Computer.

[15]  Jonathan Gagnon,et al.  Dynamic lapped texture for fluid simulations , 2016, The Visual Computer.

[16]  Hyeongseok Ko,et al.  A practical simulation of dispersed bubble flow , 2010 .

[17]  Nils Thürey,et al.  Data-driven synthesis of smoke flows with CNN-based feature descriptors , 2017, ACM Trans. Graph..

[18]  Markus H. Gross,et al.  Deep Fluids: A Generative Network for Parameterized Fluid Simulations , 2018, Comput. Graph. Forum.

[19]  Chang-Hun Kim,et al.  Efficient Representation of Detailed Foam Waves by Incorporating Projective Space , 2017, IEEE Transactions on Visualization and Computer Graphics.

[20]  Duc Quang Nguyen,et al.  Physically based modeling and animation of fire , 2002, ACM Trans. Graph..

[21]  Ronald Fedkiw,et al.  Practical animation of liquids , 2001, SIGGRAPH.

[22]  Mike Cammarano,et al.  Time Dependent Photon Mapping , 2002, Rendering Techniques.

[23]  Matthias Teschner,et al.  Eurographics/ Acm Siggraph Symposium on Computer Animation (2007) Weakly Compressible Sph for Free Surface Flows , 2022 .