分光反射特性を利用した茶の生育・品質推定手法に関する研究

Abstract

This study was conducted to develop a method to evaluate, growth and quality of tea leaves at each growing stage by spectral reflectance analysis. The spectral characteristics of tea canopies were measured using a handheld spectroradiometer during the growth season of new leaves. At the same time, the canopies were recorded with a video camera using several band pass filters. The newly picked leaves were also analyzed using an NIR analyzer to determine their chemical contents. The spectral reflectance curves were differentiated with respect to wavelength on the first and second orders. As a result, several characteristic peaks were identified. Characteristic waveband, CW was defined as the waveband of the characteristic peak ranging within 10 nm. Several indices consisting of these CW were proposed and compared with ground truth data. It was demonstrated that the normalized index, ND770, 660 = (R770-R660)/(R770+R660), strongly coincided with the results of the biomass of new leaves. Moreover, it was shown that the ND770, 540, ND770, 520, and ND540, 520 were in agreed with the quality components of the new tealeaves observed by the NIR products analyzer. Principal component analysis demonstrated that the first component changed according to a quadratic function based on time dependency. Calculating the peak value for this function, it shows that the suitable picked day is determined based on the quality of the tea leaves. AF value calculated by the ratio of amino acid and fiber content has been used as the index of the quality of tealeaves commercially. The ND710, 520 index highly correlated with the AF value. These results suggest that the spectral reflectance measured non-destructively is effective in estimating the leaf quality during the cultivating season.