Image Denoising Using Fractional Quaternion Wavelet Transform

摘要: This paper presents an image denoising algorithm using fractional quaternion wavelet transform (FrQWT). In particular, images corrupted with additive Gaussian noise are considered and FrQWT is performed via hard semi-soft thresholds. The thresholding on the coefficients reveals capabilities of in restoration degraded by noise. simple adaptive since estimation threshold parameters depends data coefficients. order captures texture details more way. Experimental results exploit better performance compared to various techniques such as discrete wavelets, complex domains terms high peak signal-to-noise ratio (PSNR).