An Intelligent Risk Detection Method for Aircraft Towing Operations Based on a Two-Stage Vision Model
摘要
To reduce the risks caused by aircraft towing operations, this study presents an intelligent risk detection framework using a two-stage vision analysis method. Firstly, TimeSformer is used to divide and identify different phases of the operation. Then YOLOv5 is applied to detect and recognize critical safety items. Tests show that the framework can accurately detect important operational points and identify possible hazards with high precision. Additionally, the method improves automated monitoring and safety in ramp operations and offers a practical new approach for intelligent aviation safety management.