연구 제목: Anomaly Detection in Industrial Pipelinese Using Thermal and RGB Images
시상일: 2025년 7월 25일
연구자: Jenipher Siriwa, Nehemiah Balozi, Eugene Auh, Hyungpil Moon
Conference Name: International Conference on Precision Engineering and Sustainable Manufacturing (PRESM2025)
Abstract : This research presents an automated anomaly detection system for industrial pipelines using combined thermal and RGB imaging. Trained with Mask R-CNN and YOLOv8, the system achieves 98% detection accuracy. The dual-vision approach enhances reliability for real-time monitoring and predictive maintenance.
