1: \begin{abstract}

2: This paper extends to two dimensions the recent signal analysis method based on the semi-classical analysis of the Schr\"odinger operator. The generalization uses the separation of variables technique when writing the eigenfunctions of the Schr\"odinger operator. The algorithm is described and the effect of some parameters on the convergence of this method are numerically studied. Some examples on image reconstruction and denosing are illustrated.\\

3: 

4:  \textit{Keywords --} Image reconstruction, image denoisng,  Schr\"odinger operator, discrete spectrum, localized functions, semi-classical analysis.

5: \end{abstract}

6: