Search

BACKGROUND Non-destructive methods based on fluorescence hyperspectral imaging (HSI) techniques were developed to detect worms on fresh-cut lettuce. The optimal wavebands for detecting the worms were investigated using the one-way ANOVA and correlation analyses. RESULTS The worm detection imaging algorithms, RSI-I(492–626)/492, provided a prediction accuracy of 99.0%. The fluorescence HSI techniques indicated that the spectral images with a pixel size of 1 × 1 mm had the best classification accuracy for worms. CONCLUSION The overall results demonstrate that fluorescence HSI techniques have the potential to detect worms on fresh-cut lettuce. In the future, we will focus on developing a multi-spectral imaging system to detect foreign substances such as worms, slugs and earthworms on fresh-cut lettuce. © 2017 Society of Chemical Industry