基于全局视角的无人车导航系统

Unmanned Vehicle Navigation System Based on Global Perspective

针对传统无人车的车载摄像头视野狭隘的问题,设计了利用高位双目摄像头提供全局信息的导航方式,通过图像的预处理,结合车载传感器等多源传感器的数据融合实现周围环境的信息处理,并且获取道路的颠簸度并进行障碍物的判断,实现无人车动态路径选择,利用MATLAB进行环境建模,验证了基于全局视角的无人车导航系统的可行性,该系统具有一定的推广应用价值。

Aiming at the narrow vision problem of on-board camera for traditional unmanned vehicle, a navigation method is designed by using high binocular camera to provide global information. The image processing is combined with the data fusion of the multi-source sensor such as the vehicle sensor to realize the information processing of the surrounding environment. The bumpiness of the road is obtained, the obstacles are judged and the dynamic path of the unmanned vehicle can be selected. By modeling the environment in MATLAB, the feasibility of the unmanned vehicle navigation system based on the global perspective is verified. And the system has certain application value.