基于Bernstein滤波器自适应提升小波变换

Bernstein Filter-Based Adaptive Lifting Wavelet Transform

该文引入了基于提升法自适应离散小波变换,探讨了空间域提升结构中预测算子和更新算子的设计,根据预测误差最小原则使伯恩斯坦预测算子自适应匹配特定的数据序列,最后由数值仿真验证了该算法的性能,对于硬域值法小波系数去噪,自实应提升小波变换同一般的小波变换相比,去噪后的信噪比效率相近,提升方法的优点在于其设计上的灵活性和计算简单,易实现自适应变换。

This paper develops adaptive wavelet transform based on the lifting scheme. The lifting construction exploits predictor and update operator design in a spatial domain. Bernstein filter predictor adaptively match a desired signal by optimizing data-based prediction error criteria. At last we discusss the performance of the algorithm by comparing the algorithm with classica wavelet transform on denoise of one-dimension testing signal according to hard thresholding of wavelet coefficient. The SNR efficiency of all method is very similar. The advantages of lifting scheme lie on its design flexibale, compute in-place and easily adaptively transform.