부산대학교 도서관

The objective of this study is to investigate the variations of optical properties during the progression of black spot disease and select the characteristic wavelengths for disease detection. The optical absorption and scattering properties were measured on winter jujubes using a single integrating sphere detection system in the band range of 900~1650 nm combined with inverse adding-doubling (IAD) algorithm in this study. The characteristic wavelengths were obtained depending on the correlation analysis between optical properties absorption and physicochemical indicators. Finally, the effectiveness of characteristic wavelengths was verified using near-infrared spectroscopy. The results indicated that a* and weight loss increased with storage time, while L*, soluble solids content and chlorophyll showed a consistent decrease. Both the absorption coefficient (μa) and the reduced scattering coefficient (μs') of the winter jujubes showed a significant downward trend with the development of black spot disease. The μa and μs' curves were highly correlated with disease extent (spot area) and some quality indicators (weight loss, L*, a*, soluble solids content and chlorophyll) in the 1400~1650 and 900~1360 nm, respectively. The discriminant models built based on the 10 characteristic variables selected by optical properties performed best compared with the models based on full wavelengths and algorithm-based characteristic wavelengths, with an overall accuracy of 92.53% and 92.35% for the calibration and prediction set, respectively. This study demonstrated the potential of the short- and mid-wave infrared optical signal in the disease detection on winter jujubes and the effectiveness of characteristic wavelengths on the basis of optical properties.