Wavelet transform in high dimensionality spectral data classification
Wei Zhou · Journal of Foshan University · 1999
Wavelet transform as a powerful theory has been extensively applied to signal processing in recent years. According to the property of wavelet transform in high demensionality spectral data classification, the paper investigated a feature extraction which could map P dimensional vectors into P′ dimensional vectors and decrease classification model parameters while simultaneously retaining the majority of discriminatory information. According to different wavelet bases, the paper used m band discrete wavelet transform coefficients for classification and presented its algorithm. The results showed that the classification rate was quite satisfactory.