Abstract A naturally system is usually modeled as a causal system, in which the present output is determined by the past inputs. In contrast, the noncausal system is modeled by the future inputs in addition to the past inputs, and is also less explored. In this thesis, we apply the noncausal modeling to the image restoration for the blurred images corrupted by additive white Gaussian noise. We applied three methods for our image deblurring problem. The first method is exploiting the compound Gauss-Markov image model, which has been proven useful in image restoration. The image is restored in two steps iteratively: restoring the line field by the assumed image field and restoring the image field by the just computed line field. The second method is to apply the Kalman filter using the above the compound Gauss-Markov image model and the line field. The third method is to apply the Kalman filter without using the line field. Our experiments have shown the second method to be the best among the three methods.