Non-Destructive Prediction of Paddy Seed Quality using Near Infrared Spectroscopy

Abstract

The need for a rapid and non-destructive method for measuring paddy seed quality is important since it could help save time and effort of the breeders and farmers. Near-infrared (NIR) spectroscopy which is a non-destructive analytical technique was used in this study. It was used to predict the germination rate and vigor of hybrid paddy seeds and classify the seeds according to the level of germination rate. A total of 71 seed lots with varying germination rate and vigor were used in the study. Paddy seeds were classified according to their germination rate (very high >90%, high 80-90%, moderate 70-79% and low <70%). Each seed lots were scanned at five random locations. Reflectance values of paddy seeds were scanned using NIR spectrometers. After scanning, the germination rate was determined using standard methods in the laboratory. Pre-processing techniques were done to the spectral data gathered. The preprocessed spectral data were subjected to principal component analysis and partial least square regression (PLSR) to predict the paddy seed quality parameters. Performance of PLSR using the PCA scores from the preprocessed spectral data of paddy seeds showed that the calibration model for predicting the germination rate had a computed R² value of 0.626-0.804 and the RMSE was 11.716-8.489. The R² and RMSE of the model for seed vigor was 0.670-0.787 and 14.095-11.328, respectively. The model for predicting the seed class had a computed R² value of 0.803 and the RMSE was 0.514 for the NIR spectrometer and the R2 and RMSE was 0.648 and 0.687, respectively, for the vis NIR spectrometer. Based on the RMSE values and using the guidelines by Williams and Norris (2001) the models for seed class can be used with caution in most applications like in research. Results showed the potential of NIR spectroscopy as a non-destructive method of predicting seed quality parameters and classifying paddy seeds according to the level of germination rate. The results of the study could be used to predict seed germination rate and vigor and classify sort seed lots quickly based on the values of germination rate and seed vigor

Citation:

FLORES, F., FOLIENTE, J., YAPTENCO, K., & ARMSTRONG, P. (2020). Non-Destructive Prediction of Paddy Seed Quality using Near Infrared Spectroscopy. Philippine Journal of Agricultural and Biosystems Engineering, 16(2), 65–74. https://doi.org/10.48196/016.02.2020.05