Lateral inhibition based holistic approach to adaptive image enhancement

We present a physiologically inspired adaptive algorithm for noise removal in an image while preserving significant amount of edge details. The algorithm is motivated by the classical lateral inhibition based receptive field in the visual system as well as the holistic approach of the well known bilateral filter. We propose an adaptive difference of Gaussian (DoG) filter with varying window size depending upon the edge strengths in the image. Our algorithm has advantages over similar other techniques such as simple Gaussian filter, DoG filter, and is comparable to the bilateral filter in terms of edge enhancement. Furthermore, time complexity of our algorithm is much less than the bilateral filter.