Improvement of the noisy signal pattern recognition using wavelet transform and weighting coefficients for each subblock

Issue title: Proceedings of the twelfth International Symposium on Interdisciplinary Electromagnetic, Mechanic and Biomedical Problems, ISEM Bad Gastein

Article type: Research Article

Authors: Midorikawa, Yoichi^{a; *} | Akita, Masanori^a

Affiliations: [a] Faculty of Engineering, Oita University, 700 Dannoharu, Oita, 870-1192 Japan

Correspondence: [*] Corresponding author. Tel.: +81 97 554 7817; Fax: +81 97 554 7817; E-mail: [email protected]

Abstract: This paper shows the method for improving of the noisy signal pattern recognition. We have proposed the method for recognition of signal polluted with noise using spectral modification method by rule and smoothing method by wavelet transform. However, the modified spectral envelopes are flattened and added valleys become ambiguous. We applied emphasis method by weighting coefficient for each subblock. As a result, spectral shapes and performance of the pattern recognition are improved.

Keywords: Signal pattern recognition, noise, cepstral analysis, wavelet transform

DOI: 10.3233/JAE-2007-770

Journal: International Journal of Applied Electromagnetics and Mechanics, vol. 25, no. 1-4, pp. 743-747, 2007