Comparison of the Fuzzy-based Wavelet Shrinkage Image Denoising Techniques
Farshad Tajeripoor, M. JavadZomorodian, Shirvan Branch · 2012
In this paper, a comparative study on the different membership functions which are used for fuzzy-based noise reduction methods is done. This study focuses on the three different membership functions such as Gaussian, Sigmaf and Trapezoidal. The fuzzy wavelet shrinkage method is tested with different membership functions in order to reduce different types of noise such as Gaussian, Salt & Pepper, Poisson and Speckle. The measure of comparison between different membership function is based on PSNR (Peak Signal to Noise Ratio). Experimental results show that on the some well-known images, such as "Lena", "Barbara " and "Baboon", the Gaussian membership function can efficiently remove the additive Gaussian and the Poisson noises from the grey level images. Furthermore, on the Speckle and Salt & Pepper noises, the Sigmaf membership function outperforms the Trapezoidal one to remove noise. Keywords:Fuzzy set, Membership function, Noise detection, Noise reduction, Wavelet shrinkage.