Spatio-temporal approach for time-varying image motion estimation

Image motion estimation using the spatio-temporal approach has largely relied on the constant velocity assumption. Spatio-temporal algorithms are developed in this paper to handle time-varying motion. Time-variation is modeled either as a polynomial or as a periodic function of time. Under these models, it is shown that time-varying image motion estimation is equivalent to parameter estimation of one-dimensional polynomial phase or phase modulated signals. Simulation results are provided to illustrate the proposed algorithm