AIG Based Nonlinear Anisotropic Smoothing Strategy for Vector-Valued Images

• Published in: Shanghai jiao tong da xue xue bao Vol. 14; no. 2; pp. 223 - 228
• Main Author: 张相芬 田蔚风 陈武凡 叶宏
• Format: Journal Article
• Language: English
• Published: Heidelberg Shanghai Jiaotong University Press 01.04.2009
• Subjects: Architecture; Computer Science; Electrical Engineering; Engineering; Life Sciences; Materials Science; 仿射不变量; 信号噪声比; 各向异性扩散; 向量值; 非线性; affine invariant gradient (AIG); affine invariant nonlinear anisotropic diffusion (AINAD); TP 391; diffusion tensor imaging (DTI); smoothing; Euclidean invariant gradient (EIG)
• ISSN: 1007-1172; 1995-8188
• DOI: 10.1007/s12204-009-0223-z

The effects of the Rician noise on the calculated tensors are analyzed and an affine invariant gradient （AIG） based nonlinear anisotropic smoothing strategy is presented. The AIG based smoothing strategy is a development of the affine invariant nonlinear anisotropic diffusion （AINAD） restoration model, introduced by Guillermo Sapiro, and adopted to restore vector-valued data. To evaluate the efficiency of the presented AINAD model in accounting for the Rician noise introduced into the vector-valued data, the peak-to-peak signal-to-noise ratio （PSNR）, signal-to-mean squared error ratio （SMSE） and Beta（parameter that stands for edge preservation） metrics are used. The experiment results acquired from the synthetic and real data prove the good performance of the presented filter.