Real time obstacle detection method based on lidar and wireless sensor

In order to improve the accuracy of laser radar for dynamic obstacle detection, and ensure the safety of autonomous vehicles, the method of dynamic obstacle method of wireless sensor and 3 dimensional laser radar Velodyne based on information fusion of vehicle identification and a multi feature detection and tracking are put forward. A preliminary model of mobile obstacle is established by using lidar, and the model information includes center position, length and width; at the same time, in order to reduce the number of sensor nodes to reduce the energy consumption, the tracking area is designed, and the moving objects are modeled the second time by using the wireless sensors' data in the tracking area, and if data consistency is lower, data fusion and model updating are performed; and the optimized Euclidean algorithm is combined with the corresponding features to complete the tracking of obstacles. The calculation results show that the method proposed in this paper can solve the problem that the accuracy of detection and modeling is not high for occlusion, and it can guarantee the driving safety of autonomous vehicles.