Lightweight Visual Detection System for Object Identification with ArUco Markers in Resource-Constrained Environments

Yulianto, Koko Edy
Saputro, Rujianto Eko
Utomo, Fandy Setyo

Electronic Journal of Education, Social Economics and Technology
Vol 6, No 2 (2025)

Object detection is a fundamental task in computer vision systems used in robotics, automation, and real-time tracking applications. However, implementing accurate and responsive detection on low-cost embedded hardware presents significant challenges due to limited processing power and environmental variability. This study aims to evaluate the performance of an object detection system utilizing ArUco markers on a Raspberry Pi-based platform. The research investigates the system’s ability to detect and identify three types of physical objects—a plastic bottle, a flower pot, and a glass cup—as well as the performance when all three objects are present simultaneously. The system was tested under controlled static conditions using a camera to capture real-time video streams. Detection time, computation time, and accuracy were measured across five consecutive frames for each scenario. Results show that the system achieved consistent detection and processing times below 0.14 seconds per frame, meeting real-time performance criteria. Detection accuracy across all individual object scenarios exceeded 91%, with the highest accuracy recorded in the multi-object scenario at 93.44%. No detection failures occurred during the experiments, and frame-by-frame analysis confirmed temporal stability. These findings indicate that marker-based detection is a reliable and efficient approach for real-time applications in structured environments. The study provides a foundation for extending the system to more dynamic conditions in future research.