A NONLINEAR MODEL FOR VELOCITY ESTIMATION FROM INFRARED IMAGE SEQUENCES

Velocity estimation using tracer conservation or heat flow equation based on infrared image sequence is one of the most challenging inverse problems in geosciences and remote sensing applications. In this paper, a new nonlinear model has been created for estimating top surface velocity from an infrared image sequence. The differential form of tracer conservation or heat flow equation is replaced by a temporal integral form of the heat flow conservative constraint equation in which the initial and final states of flow terms are associated with only two time-varying frames at time t(1) and t(2). Iterative equations with Gauss-Newton and Levenberg-Marguardt algorithms for the estimation of the top surface velocity are formulated based on the temporal integral form of the heat flow equation, modelling the velocity field, and a nonlinear least-squares model. The solution of a numerical model is used as a benchmark to exam the new estimator. Both angular and magnitude error measurements based on the synthetic surface heat flow from the numerical model demonstrate that the performance of the new approach with the nonlinear model are much better than the results of using a linear model of the tracer conservation or heat flow equation. One sequence of NOAA AVHRR images taken in the New York Bight fields is also used to demonstrate the performance of the nonlinear inverse model.