Parallel Optical Snow and Sampling Importance Re-sampling (SIR)

Parallel optical snow is a particular kind of image sequence where all of the objects move parallel to one another. The problem is to determine the direction of this parallel optical snow. It has been shown that there exists a solution to this problem via a Fourier transform. Our objective is to solve this problem without using a Fourier transform. Instead, we use Sampling Importance Re-sampling (SIR) modified to suit our particular purposes. Our results show that this is a reliable method, which can find the solution to parallel optical snow after only observing two frames of the image sequence.