The bus driving behavior in the bus station area is mainly characterized by theoretical models such as follow-up, lane change, and simulation. In order to build a behavioral model that matches the actual operation of the vehicle and clarify whether the existing bus station related design specifications are scientific and reasonable.This paper presents the method parametrically extracting real-life behavior data including trajectory, velocity, acceleration when the bus enter and exit station based on UAV and image recognition technology, which detailed analyzing the law of movement about vehicles. Furthermore, we discuss the necessity and feasibility of building a vehicle movement model based on measured big data and propose the corresponding technical route and key technologies. The preliminary research demonstrates the method can accurately describe the organic links including bus driving behavior , road design and social vehicle operation, which can provide technical support for geometric design of bus stops, vehicle control and driverless research.