Total Variation based Multivariate Shearlet Shrinkage for Image Reconstruction
Autores:
Deng, Chengzhi; Nanchang Institute of Technology Hu, Saifeng; Nanchang Institute of Technology Tian, Wei; Nanchang Institute of Technology Hu, Min; Nanchang Institute of Technology Li, Yan; Nanchang Institute of Technology Wang, Shengqian; Nanchang Institute of Technology
Fecha:
2013-01-01
Publicador:
TELKOMNIKA: Indonesian journal of electrical engineering
Shearlet as a new multidirectional and multiscale transform is optimally efficient in representing images containing edges. In this paper, a total variation based multivariate shearlet adaptive shrinkage is proposed for discontinuity-preserving image denoising. The multivariate adaptive threshold is employed to reduce the noise. Projected total variation diffusion is used to suppress the pseudo-Gibbs and shearlet-like artifacts. Numerical experiments from piecewise-smooth to textured images demonstrate that the proposed method can effectively suppress noise and nonsmooth artifacts caused by shearlet transform. Furthermore, it outperforms several existing techniques in terms of structural similarity (SSIM) index, peak signal-to-noise ratio (PSNR) and visual quality.