A new iterative scheme for waveform estimation with application to frame-to-frame image motion estimation

A new gradient-based technique for waveform estimation is developed and implemented for frame-to-frame image motion estimation. The algorithm assumes that the waveform components are sample functions from zero-mean Gaussian fields with known covariance functions. Using the steepest ascent routine, the procedure seeks the maximum a posteriori estimates of the Karhunen-Loeve expansion coefficients of the waveform. Necessary conditions are established for the convergence of the algorithm in the presence of noise. Experimental results are presented that illustrate the behavior of the algorithm for real images