Abstract: To address the issues of lengthy and inefficient passenger security checks in railway passenger transport stations, this paper proposed a safety risk assessment method for passengers' carried items in railway passenger transport stations that combines X-ray intelligent image recognition data and flammable and explosive gas detection data. It proposed an improved YOLO (You Only Look Once) v8l object detection model to improve the detection accuracy of small prohibited items and alleviate the problem of low model accuracy caused by item occlusion; used a flammable and explosive gas detector to detect in real time the flammable and explosive gases escaping from the luggage compartment, reduce the opening rate of security checks and improve the efficiency of passenger entry; established a safety risk assessment mechanism for passengers' carried items to evaluate the safety risk level of the carried items. The application situation shows that the proposed method for assessing the safety risks of passengers' carried items in railway passenger transport stations has improved the efficiency of passenger security checks and can provide reference for the efficient operation of railway passenger security check systems.