Smart Ripeness Assessment: Real-Time Image-Based Classification of Papaya Maturity Stages

Abstract

The classification of papaya maturity is vital for ensuring fruit quality, reducing waste and improving supply chain efficiency. Traditionally, farmers assess papaya ripeness manually by inspecting physical traits like colour, texture and softness, which is labour-intensive, subjective, time-consuming, and prone to errors. The CNN model improved accuracy but often lacked the speed needed for real-time use. This study proposes a real-time, automated papaya maturity classification system using the YOLOv8 model, which captures images using web-cam, extracts features and classifies papayas as mature, semi-mature, immature, or rotten. The results show that the YOLOv8 model gave better accuracy than the CNN model