Frequency domain filtering techniques of halftone images

When images are captured, scanned and transferred, a considerable amount of noise is added that corrupts the details present in the image. This noise that creeps into images can be from a variety of sources. It is imperative that filtering techniques are employed to enhance the quality of such images and in order to retrieve the original image. This can be done in two specific methodologies — spatial domain filtering and frequency domain filtering. Frequency domain filtering of images is extensively explored in this paper. Dual and notch filters work on centralized specific noise frequencies. The output from these filters is compared to the result obtained by using the mask technique, which involves manual masking of noise clusters present in the magnitude spectrum obtained by applying Fast Fourier Transform (FFT). The method of multiplying the mask (which can be any polygon) with the magnitude spectrum yields good results and the moiré patterns present in the image are eliminated. The removal of said moiré patterns has important applications in the role of image enhancement.