Image denoising by directional complex diffusion processes based on double-density dual-tree DWT

Abstract

Wavelet diffusion is a new popular image denoising method by combining the wavelet shrinkage and nonlinear diffusion. In this paper we extend the wavelet diffusion from real axis to complex domain and improve its performance. The double-density dual-tree discrete wavelet transform (DDDT-DWT) is a complex and directional transform. We propose an efficient image denoising algorithm by combining DDDT-DWT and the directional complex nonlinear diffusion process. We compute complex diffusion function based on directional complex wavelets. Experimental results show our algorithm is efficient.