光線路徑追蹤之高品質色散生成技術

The computer graphics in the past were limited to simple lighting or material reflection models to achieve interactive or real-time frame rate. Recently the rendering in many graphics applications has approached photorealistic quality. Physically-Based Rendering (PBR) methods are now commonly deployed in computer games, animations, movie special effects, and architecture visualization. A driving force of such a change is the quick evolution of the graphics processors. In particular, NVIDIA announced the GeForce RTX platform in 2018 to officially support ray tracing in the GPU hardware. The above mean the high-quality photorealistic images which could only be achieved in offline rendered animation movies may soon appear in real-time or interactive graphics applications. Even the gaming industry which used to rely on programmable shading techniques are now shifting to ray tracing or path tracing. Therefore, we dedicate our effort to the fundamental research that are based on the theorem of Monte Carlo Path Tracing, and GPU acceleration for path tracing to achieve real-time photorealistic rendering. The project is a continued effort of our project of the previous year. We work on the techniques that are based on Manifold Next Event Estimation (MNEE) and Specular Manifold Sampling (SMS), then extend their theoretic framework to consider the variation of light wavelength. We have delivered a novel algorithm that can produce high-quality caustics and color dispersion effects through refractive materials.