The rapid detection of the infected seedlings of Amorphophallus muelleri using Visible Near-Infrared spectroscopy

Abstract

Porang (Amorphophallus muelleri) reproduces vegetatively using the bulbil, harvested during complete dormancy and spontaneous petiole detachment. However, the bulbils can be infected by fungi under certain harvesting and storage conditions, giving rise to nonuniform seed quality, and decreased plant growth and crop yields. This study aimed to develop a non-destructive technique for detecting infected and non-infected bulbils of porang using visible-near infrared (VIS-NIR). Additionally, 90 samples were used and measured five times in various conditions involving a calibrated set of spectra ranging from 450 – 950 nm and a binary label as the predictor (X) and criterion (Y), respectively, due to the qualitative nature of the data. The bulbil was determined quantitatively using VIS-NIR spectroscopy involving Principal Component Analysis (PCA), Partial Least Squares (PLS), Partial Least Squares-Discriminant Analysis (PLS-DA), and various preprocessing data. A PCA model was used to accurately determine the variance using two principal components (PC), including PLS and PLS-DA models. The PLS model was used to calculate the variance with accuracy and RMSE of 97.65% and 7.67%, respectively, while the variance was calculated with 100% accuracy using the PLS-DA model on all pre -processed data. Therefore, the narrow spectrum ranging between 450 and 950 nm enabled the implementation of low-cost assays, such as visible near-infrared spectroscopy, and the rapid detection of harmful contaminants during the chemical studies of fungal-infected bulbils.