An Empirical Analysis of Frequency Domain High Pass Filters on Various Types of Noises

Enhancing the pictorial information for human interpretation is always been a challenging task in digital image processing. Preprocessing is used to remove the unwanted data in digital images. Frequency domain techniques are applied on Fast Fourier transformations of an image. High pass filters are often used to sharpen the digitized image and improve minutiae details. It is frequently applied on the fingerprint images. One of the main objectives of this work is to find the suitable high pass filter for desired noise type. Quality of the images is assessed with image quality metrics like PSNR, MSE and MAE.