Extraction of Lambert parameter from leaf scale hyperspectral images

Leaf scale hyperspectral images acquired by low altitude hyperspectral observation systems are promising data sources for precision farming. However, the leaf scale hyperspectral data are affected by shading and various illumination conditions so that it is difficult to obtain consistent characteristics of the spectral information. The aim of this paper is the extraction of Lambert coefficients as inherent leaf spectral profile. A dichromatic model-based method utilizing leaf scale hyperspectral data is proposed in order to diminish the spectral difference caused by illumination condition and BRDF. The results show that indices of chlorophyll consent based on the estimated Lambert coefficients are consistent with the growth stages of paddy field whether the illumination condition is clear sky or overcast.