一种改进的实时软阴影算法的设计与实现

在解决实时渲染时的阴影边界走样时,常用的(Percentage Closer Filter，PCF)方法不仅效率低下,而且边界处的平滑效果也很有限。改进了PCF算法,使算法的执行效率更高,且有更好的反走样效果。首先分别以相机和光源为基准来构造裁剪坐标系,通过2次绘制来锁定产生阴影的像素区域,然后用改进的高精度滤波模板对阴影进行适应性反走样处理,使平滑操作集中在对视觉效果有重要影响的边界区域上,有选择地优化了数据,从而大幅提高了全局计算效率。实验结果证明,文中算法比常规的PCF算法在执行速度上有明显提高,效果上也得到一定程度的改善,达到了效率与效果的优化折中。

Design and Implementation of An Algorithm of Real-time Shadow Rendering

In realistic graphics, how to reflect realistic shadow effect is an important aspect to enhance the sense of reality of the model. The shadow should meet the real physical situation as far as possible, meanwhile, a smooth edge is necessary. However, rectangular array of pixels will easily cause jagged aliasing phenomenon at the edges. How to deal with the aliasing problems effectively has been a hot topic in the field of graphics. When solving aliasing of the shadow boundary in real-time rendering, the conventional Percentage Closer Filter (PCF) is not only inefficient, but also the smoothing effect of the boundary is limited. This paper improves the PCF algorithm, the improvement makes the algorithm more efficient in implementation. This new algorithm is based on conventional Shadow Mapping and uses higher accuracy of the filter template to deal with aliasing adaptively, especially to the area which has impact on visual effects, .thus increasing the overall computational efficiency substantially. The experiment results show that this algorithm is obviously improved in efficiency compared with the conventional PCF.