Investigating the Relationship between the Reflected Near Infrared Light and the Internal Quality of Pineapples Using Neural Network

Abstract

One of the important internal qualities of pineapples is the total soluble solid content (SSC). Normally, the SSC can be evaluated using a reflectometer that is destructive and time-consuming. This research investigates the relationship between the reflected near infrared light and the internal quality of pineapples non-destructively. Five light emitted diodes (LEDs) that are in the range between 750 nm and 950 nm were used as the light source. The photodiode (OPT101) sensor was used to collect the light from the pineapple. The digital reflectometer was used to determine the reference SSC. The Near-infrared (NIR) data and the digital refractometer data were used to build the predictive model. The relationship between the near infrared light and the SSC of the pineapple was determined using artificial neural network predictive model. The internal quality of pineapples was determined using five NIR data wavelengths, the result points out that the k-fold cross-validation accurate classification was 75.56%. Besides, findings indicate that the artificial neural network that used four wavelengths that were 780 nm, 850 nm, 870 nm, and 940 nm achieved better classification than that used five wavelengths that included 910 nm. Thus, the artificial neural network coupled with NIR light is promising to be used to classify the internal quality of pineapples non-destructively.