Automated Classification of Crystal Guava Based on Weight and Color using Mamdani Fuzzy Logic

Smart agriculture cannot be separated from automation technology. Automated classification of crystal guava (Psidium guajava L. var. Crystal) based on weight and color plays a crucial role in ensuring fruit quality and meeting market demands. Traditional manual sorting methods are often inefficient, subjective, and prone to human error. This study proposes the implementation of a Mamdani fuzzy for the intelligent classification of crystal guava. The system utilizes weight and color as input variables to determine fruit quality categories. The Mamdani fuzzy model was developed by defining membership functions and fuzzy rules tailored to the characteristics of crystal guava. The proposed model was tested using a dataset collected from local farms, achieving high accuracy in classifying the fruits into predefined quality categories. The results demonstrate that the Mamdani fuzzy provides a robust, efficient, and scalable solution for fruit classification, significantly reducing dependence on manual sorting with accuracy of 96%. This research contributes to advancing intelligent agricultural systems and promoting innovation in post-harvest fruit handling technologies.