A novel variable-filter-band discrete wavelet transform
Zhong Zhang, Hiroshi Toda, Takashi Imamura, Testuo Miyake · 2012
In this study, a new variable-filter-band discrete wavelet transform (VBF-DWT) is proposed for detecting and extracting unsteady abnormal signals. The VBF-DWT is achieved by using the band pass wavelet set (BP-WS) and band rejection wavelet set (BR-WS) instead of traditional mother wavelet (MW). Corresponding to it, in the multi-resolution algorithm, the variable-band filters (VBFs) are used instead of traditional low and high pass filters. The VBFs are designed from the BP-WS and BR-WS, which is constructed from a real-signal MW.