Performance Analysis of a Robust Wavelet Threshold for Heavy-Tailed Noises
The interesting signal is often contaminated by heavy-tailed noise that has more outliers than Gaussian noise. A robust wavelet threshold based on the minimax description length principle is derived in the ε-contaminated normal family for maximizing the entropy. Compared with classical threshold based on Gaussian assumption, the robust threshold can eliminate the heavy-tailed noise better, even if the precise value of ε is unknown, which shows its robustness. The further experiment shows that soft threshold is more suitable than hard threshold for robust wavelet threshold technique.
Zhu Zhilin & Patrick Wang
G. F. Wei et al., "Performance Analysis of a Robust Wavelet Threshold for Heavy-Tailed Noises", Applied Mechanics and Materials, Vols. 40-41, pp. 979-984, 2011