An inverse iterative algorithm for microwave imaging based on moment method solution is presented here. The iterative scheme has been developed on constrained optimization technique and is certain to converge. Different mesh size for the model has been used here to overcome the Inverse Crime. The synthetic data at the receivers is contaminated with different percentage of noise. The ill-posedness of the problem is solved by Levenberg-Marquardt method. The algorithm is applied to synthetic data and the reconstructed image is then further enhanced through the Image enhancement technique