This paper proposes an algorithm to reconstruct a
High Resolution (HR) image from a set of blurred, warped,
undersampled, and noisy measured images. The proposed
algorithm uses the affine block-based algorithm in the maximum
likelihood (ML) estimator. It is tested using synthetic images,
where the reconstructed image can be compared with its original.
A number of experiments were performed with the proposed
algorithm to evaluate its behavior before and after noise addition
and also compared with its behavior after noise removal. The
proposed system results show that the enhancement factor is
better after noise removal than in case of no noise is additive, and
show that PSNR difference is better in comparison with the
results of another system.