Image Denoising via Modified Multiple-Step Local Wiener Filter and Quaternion Wavelet Transform

This paper proposes an image denoising algorithm via the modified multiple-step local Wiener filter in the quaternion wavelet transform domain. The multiple-step local Wiener filter has shown good performance on removing the Gaussian white noise. The quaternion wavelet transform (QWT) can provide a richer scale-space analysis when compared with wavelet transform or complex wavelet transform, and which is an efficient multiscale geometric analysis method. So the combination between them can remove noise more effectively and efficiently. Experimental results verify that the proposed method improves the denoising performance significantly and is very efficient in computation time.