Image enhancement using integer sequences based windows

Image enhancement techniques using spatial and frequency domain filters are concerned with general improvement of images for human observers. Of the two, frequency domain filters can be designed using a number of techniques, of which, windowing technique is the simplest technique to obtain such realizable frequency domain filters. In this work, we discuss the effect of applying the filter coefficients generated using integer window sequences and the compare the effect of using filters designed using integer sequences as windows over conventional classical window based filters for smoothening and sharpening of the images. The mean square error plot for low pass filtered image shows that over a certain range of threshold values the filter designed based on integer sequences outperforms the filters designed using classical (real numbers) windows. High pass filters designed using integer windows are found to sharpen the images much better than the filters designed using classical windows.