Hebbian learning based FIR filter for image restoration

Abstract

Image filtering techniques have potential applications in image processing such as image restoration and image enhancement. The potential of these filters largely depends on the apriori knowledge about the type of noise corrupting the images. This makes the standard filters to be application specific. The widely used proximity based filters help in removing the noise by over-smoothing the edges. On the other hand, sharpening filters enhance the high frequency details making the image non-smooth. In this paper, we have introduced a new finite impulse response (FIR) filter for image restoration where, the filter undergoes a learning procedure. The FIR filter coefficients are adaptively updated based on correlated Hebbian learning. This algorithm exploits the inter pixel correlation in the form of Hebbian learning and hence performs optimal smoothening of the noisy images. The proposed filter uses an iterative process for efficient learning from the neighborhood pixels. Evaluation result shows that the proposed FIR filter is an efficient filter compared to average and Wiener filters for image restoration applications