Rapid Phenotyping of Chilling Stress of Paprika Leaves using Hyperspectral Short Wave Infrared Imaging

The development of climate-resistant crops is one of the alternatives methods for future food security. Traditional breeding methods have been required long period of time and intensive labors because it is necessary to implement repeated over many generations. To expedite the breeding process high-throughput screening (HTS) technique based on phenomics has been emerged recently. This technique has advantages that it can measure the plant phenotype fast and nondestructively. The objective of this study was to investigate the feasibility of hyperspectral short wave infrared (SWIR) imaging for the measurement of stress response of the entire area of paprika leaves exposed to chilling condition (5°C). The advantage of hyperspectral imaging is that it visualizes the physico-chemical response of the entire area of the sample instead of providing an averaged indicator. In order to investigate the significant spectral wavebands as indexes of chilling stress response of paprika leaves, analysis of variance (ANOVA) for entire combination of two wavebands ratio have been used. The ratio of two spectral wavebands (1448 nm and 1545 nm) has potential for evaluation of chilling stresses in paprika leaves. It is presumed that chilling stress affects the changes of chlorophyll and water contents in the paprika leaves. The classification accuracies for discriminating chilling stressed leaves from normal ones were above 95% with the best two wavebands ratio in both of the calibration and validation. The results show that hyperspectral SWIR imaging has potential for the use of nondestructive phenotyping tool for the measurement of chilling stress of the entire area of vegetable leaves.