Modulation Transfer Function Compensation through a modified Wiener Filter for spatial image quality improvement

The knowledge of Modulation Transfer Function (MTF) for a degraded image provides a mean to compensate for the image degradation, which improves the image quality in terms of sharpness. MTF compensating is in the image restoration; hence it is fundamentally an ill-posed problem. This paper proposes a stable and flexible filtering technique that executes an optimal tradeoff between sharpness and noise to warrant an acceptable result of image restoration. The MTF compensating is performed through a modified Wiener filter, and can be reduced to a well-posed problem by incorporating the regularization method. The modified Wiener filter employed the L-curve method for selection of optimal regularization parameter, and introduces an inverse control and smooth control parameter that allows freedom of tuning when the importance of image sharpness versus noise is being compromised. The data sets used in the analysis were synthetically blurred remotely-sensed images simulated from Level-2A product of IKONOS. The results by modified Wiener filter were analyzed and presented, they were found to be able to reduce the Mean Square Error by more than 50% with respect to the original image, and it results in a significant improvement of spatial image quality. Key-Words: Modulation Transfer Function, Spatial image quality, Wiener filter, Image restoration, Ill-posed problem, L-curve