Development of a Computer Vision System for the Non-Destructive Detection of Bananas

Quality assessment in fresh fruits is essential to ensure their commercial value and reduce losses throughout the supply chain. This paper presents the development of a computer vision–based system for the non-destructive detection of bananas, aiming to automate the classification process according to ripeness. The YOLOv11 algorithm was trained with a dataset consisting of 824 banana images in different states (green, ripe, and overripe). The system was implemented on a conveyor belt, incorporating 3D-printed components and an automatic segregation mechanism. Experimental tests achieved a classification accuracy of 97.2%, validating the applicability of the proposed system in agro-industrial environments.

• المؤلف الرئيسي: Durán-Fonseca, Miguel (author)
• مؤلفون آخرون: Padilla-Ayala , Jesús (author), Gudiño-Lau, Jorge (author), Charre-Ibarra, Saida (author), Alcalá-Rodríguez, Janeth (author)
• التنسيق: article
• اللغة: spa
• منشور في: 2026
• الموضوعات: visión artificial; clasificación de frutas; inteligencia artificial; inspección no destructiva; computer vision; fruit classification; artificial intelligence; non-destructive inspection
• الوصول للمادة أونلاين: https://revistadigital.uce.edu.ec/index.php/INGENIO/article/view/8351