Detection of insect-damaged vegetable soybeans using hyperspectral transmittance image Original Research Article

Insects in vegetable soybean products pose potential hazard to consumers, thus making the food industry liable for economic losses. The objective of the current study is to develop a hyperspectral imaging technique for detecting insect-damaged vegetable soybeans. Hyperspectral transmission images were acquired from normal and insect-damaged vegetable soybeans over the spectral region between 400 nm and 1000 nm for 100 vegetable soybean pods (225 beans). Four statistical image features (minimum, maximum, mean, and standard deviation) were extracted from the images for classification and given as input to a discriminant classifier. The support vector data description (SVDD) classifier achieved 100% calibration accuracy. SVDD achieved 97.3% and 87.5% accuracies for normal and insect-damaged samples, respectively, with a 95.6% overall classification accuracy, for the investigated independent test samples. Therefore, the hyperspectral transmittance technique can discriminate insect-damaged vegetable soybeans.