Discrimination of "Hayward" Kiwifruits Treated with Forchlorfenuron at Different Concentrations Using Hyperspectral Imaging Technology

To investigate the feasibility of using hyperspectral imaging technology to discriminate kiwifruits treated with forchlorfenuron at different concentrations, "Hayward" kiwifruits were treated with forchlorfenuron solutions at concentrations of 2.5, 5.0, 7.5, 10.0, and 12.5 mg/L at 15 days after petal fall. Hyperspectral images of 400 samples, including 100 untreated kiwifruits and 60 treated kiwifruits at each forchlorfenuron concentration, were acquired. Mask was done for each image and mean spectrum was calculated for analysis. The original spectra were preprocessed by standard normal variate and all samples were divided into calibration set and prediction set based on Duplex algorithm according to the ratio of 2:1. The calibration set included 67 untreated kiwifruits and 200 treated kiwifruits, including 40 kiwifruits at each forchlorfenuron concentration. Other 133 kiwifruits were used as the prediction set. Twenty-nine characteristic wavelengths were selected from the full spectra from 928.19 to 1,658.19 nm using successive projection algorithm (SPA). Four discrimination models, i.e., partial least squares discriminant analysis, support vector machine, extreme learning machine, and random forests were established with the full spectra and the selected characteristic wavelengths as inputs, respectively. The results showed that the best model, with 97.7 % accuracy rate, for "Hayward" kiwifruits treated with forchlorfenuron was SVM established using selected characteristic wavelengths by SPA. This study indicates that hyperspectral imaging technology can be used to nondestructively and accurately discriminate forchlorfenuron-treated kiwifruits with different concentrations from untreated ones.