Non-Destructive Classification Model of Pineapple Sweetness Level

Abstract

This research focuses on the development of a non-destructive classification model for determining the sweetness level of pineapples using Internet of Things (IoT) technology and the K-Nearest Neighbor (KNN) algorithm. The main goal of this study is to assess the sweetness of pineapples without damaging the fruit, which is a common issue in traditional sweetness measurement methods, such as using a refractometer that requires pineapple juice. Subang Regency was selected as the sampling location because it is one of the largest pineapple-producing areas in West Java. There are three primary markets for pineapple distribution in the region: supermarkets, processing industries, and traditional markets, each with different sweetness standards.A total of 500 pineapples were used in this study, with 450 samples used for training data and 50 samples for testing data. The research utilized a TCS230 color sensor connected to an Arduino Uno R3 to capture RGB data from the pineapple skin at a specific distance. The RGB values obtained were then processed using the KNN algorithm to predict the sweetness level of pineapples based on Brix categories: high (14-17 Brix), medium (10-13 Brix), and low (<10 Brix).After acquiring the RGB values, the data was processed to produce a valid classification of the pineapple's sweetness level. Testing was conducted using a confusion matrix, which provided metrics such as accuracy, precision, and recall. The results showed that the developed KNN model successfully classified pineapples with an accuracy of 72%, meaning that the majority of predictions made by the model matched the actual sweetness level of the pineapples.This non-destructive approach offers significant advantages over traditional methods, as it not only preserves the integrity of the pineapple but also improves efficiency in terms of time and resources. The use of IoT technology and machine learning, such as KNN, has proven to provide a more practical and accurate solution for farmers and industry players in maintaining pineapple quality and enhancing the efficiency of distribution to various markets.