The soft-thresholding and the hard-thresholding method to estimate wavelet coefficients in wavelet threshold denoising are firstly discussed. To avoid the discontinuity in the hard-thresholding and biased estimation in the soft-thresholding, three novel models of threshold estimator are presented, which are polynomial interpolating thresholding method, compromising method of hard- and soft-thresholding and modulus square thresholding method respectively. They all overcome the disadvantages of the hard- and soft-thresholding method. Finally, an example is given and the experimental results showt hat the improved techniques presented in this paper are efficient.
CITATION STYLE
Guoxiang, S., & Ruizhen, Z. (2001). Three novel models of threshold estimator for wavelet coefficients. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 2251, pp. 145–150). Springer Verlag. https://doi.org/10.1007/3-540-45333-4_19
Mendeley helps you to discover research relevant for your work.