DSRC And Sensor Fusion Based Target Tracking Method For Intelligent Vehicle Sensor Fusion

Intelligent vehicle tracking system is used for real-time tracking to the target, the target's accurate state information is the systems' necessary preconditions which are secure warning, driver assistance and automatic driving. In the cooperative intelligent vehicle scene, the traditional vehicle tracking methods rely solely on on-board sensors to track the vehicle, ignoring the information of cooperative target obtained by inter vehicle communication, there are some problems on low accuracy of data association and tracking filter. So, it is significant that researching target tracking method which used the cooperative target's information obtained by inter vehicle communication.In this thesis, aiming at the shortage of existing method for vehicle tracking in the cooperative intelligent vehicle scene, an intelligent vehicle tracking method based on the fusing DSRC(Dedicated Short Range Communications) and on-board sensors is proposed. This method is utilized a part of information that the cooperative target vehicle can publish their own status and identity information by DSRC communication technology to modify the result of data association and tracking filter, so that improving the accuracy of tracking effectively. The basic idea of this method: Firstly, the main vehicle collects data by the laser radar and process data initially, relating data of the LIDAR by data association algorithm. Secondly, receiving and processing the identity and status information of cooperative target vehicle by DSRC communication technologies. Then, the results of data association are modified with the information. Finally, track and filter the modifying result of data associate and fuse with the target vehicle state information. The effectiveness of this method is verified by comparing the simulation results.This thesis will apply the intelligent vehicle tracking method to cooperative intelligent vehicle scene, and the intelligent vehicle tracking system is designed and developed. The system contains data collection and processing module, target tracking module. The first module is responsible for all on-board sensors to collect and process the data, after processed, the data is transferred to the target tracking module for target tracking, and finally, the tracking results is displayed on the interactive interface. In this thesis, the intelligent vehicle tracking system designed and implemented according to the idea of the high cohesion and low coupling to modularize the software, so that the system has good scalability.Finally, the intelligent vehicle tracking system developed in this thesis is used in a real vehicular experiment, experimental results verify the effectiveness of the proposed algorithm and system.