Simultaneous motion parameter estimation and image segmentation using the EM algorithm

We propose and investigate a method to simultaneously segment an image in a sequence of images into regions of uncovered background, moving, and stationary pixels and estimate parameters of interest for the various regions. The basis of our method is the expectation-maximization (EM) algorithm for maximum-likelihood estimation. We view the intensity difference between image frames as the incomplete data and the intensity difference with the region identifier as the complete data.