Résumé : Data Assimilation is a well-known mathematical technic used, in environmental sciences, to improve, thanks to observation data, the forecasts obtained by meteorological, oceanographic or air quality simulation models. It aims to solve the evolution equations, describing the dynamics of the state variables, and an observation equation, linking at each space-time location the state vector and the observations. Data Assimilation allows to get a better knowledge of the actual system's state, named the reference. In this article, we first describe various strategies that can be applied in the framework of variational data assimilation to study various image processing issues. Second, we detail the mathematical setting and the analysis of pros and cons of each strategy for the issue of motion estimation. Last, results are provided on synthetic images and satellite acquisitions.