Wavelet-Based Feature Extraction for Retrieval of Photosynthetic Pigment Concentrations from Hyperspectral Reflectance

With recent advancement in precision pasture, the need for quickly assessing photosynthetic pigment content of pasture has become apparent. Hyper-spectrum technology can provide an non-destructively methods for evaluate the vegetation photosynthetic pigment content. This paper is devoted to illustrating the potential for wavelet analysis of hyperspectral reflectance signals in the field of estimating photosynthetic pigments and evaluating grass quality. Wavelet entropy and determination coefficient were used to evaluate the prediction of pigment content, and compare it with original reflectance. The results demonstrated that by decomposing canopy spectra, the resultant wavelet coefficients of reflectance around scale 32 can be used to generate accurate predictions of chlorophyll concentration, despite wide variations in the range of other biochemical and biophysical factors that influence leaf reflectance. Wavelet analysis hold the promise for extracting absorption feature accurately from hyperspectral reflectance and its derivatives using the wavelet entropy at different scales.