关于图像反走样的理论及应用

About Image Anti-aliasing Theory and Application

图像反走样是用于提升图像质量的一个重要环节。导致图像失真的主要原因是由于采样频率过低及重建过滤器的质量低下。在实时渲染的应用中,随着硬件的发展及渲染架构的改变,主流反走样技术逐渐从硬件加速模式发展到后处理模式。纯后处理算法由于缺失子像素信息,无法生成高质量的反走样图像。针对这一问题,混合反走样在保留了后处理反走样效率的同时,通过结合多重采样能提供更精确的边缘处理,并保证图像在时序上的稳定性。

Anti-aliasing plays an important role in image quality. Aliasing is mainly caused by under sampling and poor reconstruction filters. In real-time rendering application, post-process anti-aliasing becomes the mainstream technique over hardware accelerated one along with the Development of hardware and rendering architecture change. Pure post-process al- gorithm cannot generate high quality anti-aliased image as lacking sub-pixel information. To tackle this problem, hybrid anti-aliasing technique combines multi-sampling to provide more accurate edge process, and as well as temporal stability of the image, while retaining the effectiveness from post-process anti-aliasing.