The increasing volume of electronic waste (e-waste) presents challenges in efficient component recovery, particularly in developing countries such as the Philippines. This study proposes an automated system for Through-Hole Technology (THT) component sorting and fault detection to address the limitations of manual visual inspection and multimeter-based testing. The system integrates a Raspberry Pi 5 and an Arducam 64MP Pi Hawk-eye camera to perform image-based component recognition using YOLOv8 and OpenCV for identifying resistors, capacitors, and LEDs. An ATmega328P microcontroller manages dedicated test circuits for electrical validation of component functionality. Experimental results show identification accuracies of 100% for LEDs, 82.5% for resistors, and 65–75% for capacitors, while fault detection achieved 93% accuracy for resistors and 90% for LEDs, comparable to manual multimeter testing. The proposed system also doubled sorting speed for small batches. Although limitations were observed in capacitor polarity detection, the modular design demonstrates a scalable framework for improving automated component recovery and supporting sustainable e-waste management.

Keywords: Through-Hole Technology (THT), Image Processing, YOLOv8, OpenCV, Fault Detection