According to utilizing the low latency of 5G communication, a 5G mobile phone based pedestrian detection method was proposed to solve the problem that the recognition and detection of pedestrians through cameras at intersections are susceptible to weather and that the statistical range of cameras is limited.According to the four typical motion characteristics of 5G mobile terminal carried by pedestrians, such as position, speed, density and direction, this method establishes a model on the roadside equipment for pedestrian detection to reduce the error caused by pedestrian detection of a single motion feature and improve the accuracy.The validity of the method is verified by building an Internet of vehicles test platform. The experimental results show that this method can accurately count the number of pedestrians in road intersections, and work well in detection speed, which provides technical support for future research and application of intelligent transportation and safety early warning.