“…UAVs, including DJI, 3D Robotics solo and Ebee, were equipped with spectral cameras to detect frost damage of crops such as wheat( Guo et al., 2014 ; Wang et al., 2014 ; Murphy et al., 2020 ), maize( Choudhury et al., 2019 ; Goswami et al., 2019 ; Shu et al., 2022 ), oat( Macedo-Cruz et al., 2011 ), oilseed rape( She et al., 2015 ), and coffee plants( Marin et al., 2021 ; Marin et al., 2022 ). In these studies, data analysis techniques such as Pearson correlation analysis and principal component analysis were used to extract stress-related spectral features including the reflectance of different spectral bands and several commonly used vegetation indexes (VIs) such as normalized difference vegetation index (NDVI)( She et al., 2015 ), green NDVI, photochemical reflectance index (PRI), carotenoid reflectance index(CRI), and anthocyanin reflectance index(ARI)( Choudhury et al., 2019 ; Marin et al., 2021 ; Marin et al., 2022 ). Pixel-based classification with thresholds, random forest, random committee, support vector machine (SVM) and other classification methods were used to predict frost damage degree in some of these studies( Goswami et al., 2019 ; Jełowicki et al., 2020 ).…”