Banana Ripeness Detection and Servings Recommendation System using Artificial Intelligence Techniques

Abstract

The banana plantation industry is the second most widely cultivated fruit in Malaysia. The current method to detect the ripeness of banana is using chemicals in order to obtain the characteristic of the fruit. This method will harm the fruit and also affects its quality. There are also methods which are non-destructive to the products which use manpower to identify the banana. This method is time-consuming and prone to human error. Thus, the objective of this research is to determine the correct method in assisting users in selecting bananas. This project is proposing a combination of two Artificial Intelligence (AI) techniques, namely Image Processing technique and Fuzzy Logic rules in a knowledge-based system for the solution. Several samples of un-ripe, ripe and over-ripe banana are taken to identify their red, green and blue (RGB) value. The values are then analyzed to create the membership functions for the fuzzy logic. Then a set of fuzzy knowledgebased rules are implemented for the system to determine the ripeness level of the banana. From the result of the ripeness, the system will give recommendations on the fruit including suggested meal preparation and the best before the date to consume the banana. The proposed system contributes to both farmers and customers. As for the farmer, they can pick their best product to be sold to the market. While for the customers, they can choose efficiently their desired banana ripeness by using this system.