Hard mung bean seeds pose a problem in the sprouting process as they develop mold and infect neighboring seeds. Near-infrared hyperspectral imaging combined with partial least squares discriminant analysis was applied to develop a classifying model to separate hard mung beans from normal ones. The orientation of the measured beans was found to affect the classification result. The optimal partial least squares discriminant analysis model based on all orientations resulted in a correlation coefficient (R) of 0.919 with a root mean squared error of prediction of 0.197. The nongerminative parts were mapped and were concentrated at one end of the bean. Finally, a germinability index was proposed according to the proportion of colored areas between the germinative and non-germinative parts from the hyperspectral imaging results. ARTICLE HISTORY

Near-infrared spectroscopy (NIRS) in the range 900-1700 nm was performed to develop a classifying model for dead seeds of mung bean using single kernel measurements. The use of the combination of transmission-absorption spectra and re°ection-absorption spectra was determined to yield a better classi¯cation performance (87.88%) than the use of only transmissionabsorption spectra (81.31%). The e®ect of the orientation of the mung bean with respect to the light source on its absorbance was investigated. The results showed that hilum-down orientation exhibited the highest absorbance compared to the hilum-up and hilum-parallel-to-ground orientations. We subsequently examined the spectral information related to the seed orientation by developing a classifying model for seed orientation. The wavelengths associated with classication based on seed orientation were obtained. Finally, we determined that the re-developed classifying model excluding the wavelengths related to the seed orientation a®orded better accuracy (89.39%) than that using the entire wavelength range (87.88%).

Maturity classification and prediction models were built using discriminant analysis and partial least squares regression, respectively. The best classification was achieved by the model incorporating both surface visible reflectance and the resonant frequency compared with the model based on only surface visible reflectance.

No abstract

No abstract