Research on the Application of YOLOv8-based Intelligent Trolley in Natural Gas Pipeline Crack Detection

With the widespread use of natural gas pipelines, timely crack detection has become particularly important to ensure safe operation and prevent potential accidents. In response to this need, this paper designs and implements a smart car system based on the YOLOv8 object detection algorithm, specifically for detecting cracks in natural gas pipelines. The system hardware includes a smart car, ultrasonic sensors, and a high-definition camera, while the software leverages the YOLOv8 model for crack recognition. A novel multi-modal data synergy mechanism (combining visual and ultrasonic data) is proposed to effectively improve the accuracy and robustness of crack detection. Ultrasonic sensors enable obstacle avoidance, ensuring safe navigation of the car in complex pipeline environments. For crack detection, the system adopts different handling strategies based on the size of the cracks: large cracks are tracked and monitored, while small cracks are simply counted. Experimental results show that the system achieves high crack detection accuracy and reliable obstacle avoidance, effectively enhancing the automation and efficiency of natural gas pipeline inspections.