Motion Compensated Noise Reduction
Brian K. Guenter · 1994
Motion Compensated Noise Reduction Research Paper Regular Paper Stochastic sampling has proven to be an effective means of reducing many types of aliasing artifacts. However stochastic sampling requires many samples in order to produce images which appear free of noise, especially when used to antialias procedural textures or the shadowing effects of area light sources. We have developed a new algorithm that improves the signal to noise ratio of stochastically sampled image sequences. The algorithm predicts the motion of pixels in screen space, compensating out the effect of motion in the scene. Several frames of motion compensated pixel intensity information are averaged with a corresponding improvement in signal to noise ratio. The new algorithm is efficiently applied as a post process to previously rendered images. Computation time is low compared to the time required to render complex images - only 2 minutes per frame for a 320 by 243 pixel image on an SGI Indigo 2 workstation. 1. ...