在计算机图形学与虚拟现实技术领域中,自然现象的模拟作为一个难点得到了广泛的关注和研究.液体是一种最为常见的自然现象,对此类现象的建模具有极大的应用价值.传统的液体建模方法是基于物理的建模方法,但计算量较大.随着相机等采集设备的发展,出现了基于采集数据的液体建模方法.回顾了近年来国内外在液体采集和建模仿真方面的发展情况和研究进展;分析并总结了针对不同研究对象所采用的建模方法;最后提出液体采集与建模技术研究中存在的挑战及未来发展趋势.
The simulation of natural phenomena attracts a spurt of research attention and interest in computer graphics and virtual reality domain. The modeling and simulation for liquid, which isone of the most common nature phenomena, has extraordinary value in many applications like game, medical use, movies, etc. As traditional liquid modeling method, physics-based approaches suffer from the high computational complexity. Following the development of visual acquisition equipments like high- resolution digital cameras, liquid modeling method based on acquisition data has become another active research topic in virtual reality community. With this method, more realistic phenomena can be generated with technologies in computer graphics and computer vision. In addition, the visual features and rich details are preserved while the computational complexity can be significantly reduced. However, the time-varying modeling for liquid dynamics through acquisition methods is still faced with many challenges. Current research and development progress of liquid visual acquisition and simulation in recent years are reviewed and introduced. The main methods adopted in this research