@article{M964669F4,
title = "UAV Propeller Defect Location Detection System Using Sound-Based Deep Learning",
journal = "The Journal of Korean Institute of Communications and Information Sciences",
year = "2025",
issn = "1226-4717",
doi = "10.7840/kics.2025.50.6.959",
author = "Jun-Hyuk Woo, Soo-Young Shin",
keywords = "Deep Learning, Data Augmentation, UAV, Defect Detection, Audio Processing",
abstract = "With the increasing use of UAVs across various fields, their range of applications continues to expand. In particular, the utilization of UAVs in hazardous environments and urban areas has highlighted the critical importance of UAV safety. One of the primary causes of UAV accidents is propeller failure, and extensive research is being conducted to prevent and mitigate such incidents. However, with the growing demand for multi-copters designed for complex missions, identifying which specific propeller has failed among multiple propellers has become a significant challenge. This paper proposes a sound based deep learning system for classifying the location of UAV propeller failures. By utilizing a multi-channel directional microphone, the system detects noise generated from UAV propellers and employs a deep learning model to accurately the faulty propeller. This approach contributes to enhancing the reliability and operational stability of UAVs while improving maintenance efficiency. The Paper compares analysis of audio recognition results and concludes based on it."
}