Abstract: In this study, we develop and compare two different approaches for estimating the field of displacement motion. The first approach is based on the estimation of simple movement parameters (translation, rotation). These parameters are calculated by the phase method using the inverse Fourier transform or the Hough transform. The second approach estimates the global parameters of motion (Translation, Rotation, Deformation, Scaling) by using the affine model (simple and multi-resolution). These approaches are applied to the estimation of displacement fields of synthetic and real image sequences. In addition, a representation by parametric images is used in order to better characterize moving objects in a sequence of images.
M. Boughazi , N. Bonnet , M. Bedda and L. Zouaoui , 2005. Improvement and Comparative Study of two Methods for the Estimation of the Optical Flow . Asian Journal of Information Technology, 4: 694-703.