A 2-d adaptive nonlinear filter for cancellation of artifact disturbance in images

The technique of adaptive line-cancellation is successfully used in suppressing a narrow band noise in a 1-D signal. This technique can also be used in the 2-D case, i.e., in the restoration of images from an artifact disturbance. However, when the noise consists of many harmonic components, the commonly used linear filtering technique is powerless. This paper concerns a generalized nonlinear scheme of the 2-D adaptive line-cancellation, to overcome the weakness of the linear technique. A 2-D adaptive nonlinear filter termed the 2-D adaptive CPWQ filter is introduced. A simulation example is presented to demonstrate the meaning of the proposed approach in the cancellation of a hamonic-rich noise in images.