Cascaded Binomial Filter Algorithms for Noisy FTIR Spectra
Yessi Jusman, Ashidi Mat Isa · 2011
In this paper, a new algorithm has been developed for denoising a spectrum using cascade form filter. The algorithm is based on cascading binomial filter with other direct form filters to increase signal to noise ratio (S/N) of the spectrum. The proposed algorithm consists of three types namely, Cascaded Binomial-Savizky Golay (CBSG), Cascaded Binomial-Triangular (CBT), and Cascaded Binomial-Rectangular (CBR) filters. Different levels of additive white gaussian noises (AWGN) are added to cervical cell FTIR spectra (i.e. as case study) and processed it using the proposed algorithms. The algorithms are proven to minimize the effects of the AWGN and increased S/N as compared to other compared smoothing filters.