Fuzzy Notch Filter for Periodic and Quasi-Periodic Noise Reduction in Digital Images
Najmeh Alibabaie, Alimohammad Latif · SSRN Electronic Journal · 2019
Periodic noises damage the visual quality of images by imposing repetitive patterns on them. In the Fourier amplitude spectrum, they appear as spike-like components. In this research work, we introduce a new method that is based on fuzzy systems for de-noising periodic noise. The position of a frequency coefficient in the origin shifted Fourier transformed image and the current coefficient’s amplitude to a local median ratio are used as inputs of the fuzzy system. The fuzzy system detects the approximate position of the peaks in the spectrum amplitude by these inputs. The output of the fuzzy system will be a restoration mask. It is easy to be used for noise removal. Another advantage of the approach is keeping the low-frequency region of the ordinary image and preventing any change to the DC information of the image as noise. We implemented the proposed method and evaluated its performance against some images corrupted with periodic noise. The experimental results show an acceptable level of performance. Overall, this research implies that the procedure conducted by experts in the notch filter can be automated by using a fuzzy system.