Classification of Mango Species and Ripeness Using Feature Extraction with an Artificial Neural Network

• Published in: Lecture Notes in Electrical Engineering
• Main Author: Hong K.L.; Sariff N.; Ismail Z.H.; Algeelani N.A.; Sooriamoorthy D.; Mat Isa N.
• Format: Conference paper
• Language: English
• Published: Springer Science and Business Media Deutschland GmbH 2024
• Online Access: https://www.scopus.com/inward/record.uri?eid=2-s2.0-85205340303&doi=10.1007%2f978-981-97-2007-1_17&partnerID=40&md5=7bf09102b7defd5a91c6ee3fdb5dbae9

The mango is a popular fruit with over 57 million metric tons produced worldwide. With the rise of artificial intelligence and assistive robotics, this study aims to develop a model and prototype for identifying mango species and their ripeness using a Convolutional Neural Network (CNN), targeting a minimum accuracy of 90%. The research seeks to contribute to agricultural practices and assist consumers, particularly those with visual impairments. The methodology involves capturing images of mangoes, processing them through the YOLOv8 object detection model, and analyzing them with CNN algorithms. The results indicate that the system can successfully differentiate between mango species and their ripeness with a 91.67% accuracy rate. Future work could expand the variety of mangoes and stages of ripeness for a more comprehensive application. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024.