科研之友
微信
新浪微博
Facebook
分享链接
摘要
By analyzing characteristics of wavelet-based image threshold denoising, a biorthogonal wavelet of even symmetry at the zero point with (13-3) filters length and 2/4/6-order vanishing moments is constructed using a filter parameterization method. In light of the disadvantages of global threshold, the self-adaptive hierarchical threshold denoising algorithm is proposed, where the noise decay rate in detail coefficients (detcoef, for short) of wavelet decomposition was employed to calculate hierarchical threshold value. The simulation test verifies that the constructed wavelet has favorable denoising capacity such that image details can be preserved more completely. When combined with the self-adaptive hierarchical threshold denoising algorithm, the wavelet can improve image quality and SNR significantly.
