Low-Cost, Multisensor Nondestructive Banana Ripeness Estimation Using Machine Learning

This project presents a low-cost banana ripeness estimation system using Raspberry Pi, a camera, LCD display, MQ135 gas sensor, temperature sensor, pH sensor, relay module, and DC water pump. The system uses a YOLOv8-based machine learning model to analyze banana images and classify them into four categories: unripe, ripe, overripe, and rotten. The MQ135, temperature, and pH sensors provide additional information related to the ripening process, improving classification accuracy. The results are displayed on the LCD, enabling real-time and nondestructive fruit quality assessment. The system helps reduce food waste and supports efficient fruit handling and storage.

Keywords

Raspberry Pi, YOLOv8, Banana Ripeness Detection, Machine Learning, Computer Vision, MQ135 Sensor, Fruit Quality Monitoring.