Research on Fire and Smoke Detection Technology in Oil and Gas Stations Based on YOLOv10

With the rapid development of the global oil and gas industry, oil and gas stations serve as critical nodes for the extraction, transportation, and storage of petroleum and natural gas, undertaking essential production tasks. However, these stations are filled with flammable and explosive gases and equipment, making fire incidents potentially catastrophic, leading to severe property damage and casualties. Traditional fire and smoke detection technologies face numerous challenges in the complex and hazardous environments of oil and gas stations, particularly in the early stages of a fire, where detection system response time and accuracy are crucial. In recent years, deep learning-based object detection algorithms, especially the YOLOv10 (You Only Look Once) algorithm, have gained attention due to their efficiency, accuracy, and real-time performance, offering new approaches for fire and smoke detection research. This paper proposes a fire and smoke detection method for oil and gas stations based on the YOLOv10 algorithm and validates its effectiveness and feasibility through experiments. Experimental results demonstrate that the YOLOv10-based fire and smoke detection system can achieve efficient and precise fire detection in complex environments, with an accuracy of 97.2%, and respond within a short time, significantly enhancing the safety assurance of oil and gas stations.