Publications

Digital image processing includes an important approach related to image filtering that leads to image enhancement. In general, Images may get noisy according to various reasons, in such situations; the filtering approach of images is the key for smoothing and sharpening process. In this paper, we proposed a filtering technique related to the frequency domain for both lowpass and highpass filters. The idea of the proposed filter can be described in two cases, in case of lowpass filter, we set the values of the bright objects in the spectrum to ones, and the rest of them to zeros, in case of highpass filter, we set the values of the bright object in the spectrum to zeros, and the rest of them to ones. The proposed technique can estimates the perfect match on the image spectrum resulting from Fourier Transform. This technique is to cover the intensity levels starting from the initial coordinates at the center toward the various directions of the spectrum. The Intensity levels appear in various distributions and it depend on the transformed image spectrum. We can see the image spectrum as an intensity image in a spatial domain, we want to segment it in order to detect the bright objects or the low frequency in the spectrum. We have segmented the image using Otsu method, which used to define the location of low frequencies. The segmented image will be used as a lowpass and highpass filters.