Fruit Fly Detection and Classification in IoT Setup

In the 21st century, advancements in hardware and processing speeds have enabled the deployment of machine learning algorithms such as YOLOv5 on Internet of Things (IoT) devices. These AI-based IoT devices are powerful computers suitable for embedded applications, AI IoT, and edge computing, delivering the performance of a modern AI module at a reduced cost. This conference paper presents a modified version of YOLOv5 called YOLOv5-FlyEye, optimized for embedded systems such as the Nvidia Jetson Nano. The proposed modification achieves high performance in detecting and classifying fruit flies by reducing the computational complexity of the YOLOv5 algorithm. The study aims to demonstrate the potential of AI-based IoT devices in the field of agriculture, particularly in pest detection and monitoring, and contributes to the development of optimized algorithms for efficient and cost-effective embedded systems.