Digital fuzzy random impulse noise reduction for grayscale images

Abstract

Fuzzy Logic (FL) based algorithms have emerged as one of the most interesting areas for many applications. The high growth of fuzzy logic applications led to the need of finding efficient way for its hardware implementation. Therefore, the implementation of a FL through the use of the Very high speed integrated circuits Hardware Description Language (VHDL) code is also becoming popular. In this paper, a novel FL based filtering technique is used to restore the images corrupted by impulse noise i.e., detection of impulse noise from the digital images. Theis scheme is in two ways namely the detection of noisy pixels and process the corrupted pixels to remove the noise which are implemented as (i) center pixel of the window is tested whether impulse noise is present or not, (ii) replace the noisy pixels by using FL algorithm and (iii) create a histogram if the image and again remove the noise. The performance of the proposed technique is tested on various test images corrupted at various noise densities. It is found that the proposed technique for the detection and removal of impulse noise from the gray scale images is very good rather than exiting techniques.