Automatic Driving Multi-target Tracking Motion IMU Local Information Com-pensation Optimization
Multi-target tracking is one of the key modules in the automatic driving system,and its result mainly de-pends on the accuracy of the data association process in the tracking module.By introducing the inertial measure-ment unit(IMU)or global positioning system(GPS)data of the observation point itself,the rotation and displacement relations between the local coordinate system of the current frame and the local coordinate system of the previous frame are calculated after the arrival of a frame of data,and the motion compensation is carried out for the state of the tracked object according to the obtained coordinate transformation relationship.Make it offset the offset caused by the motion of the observation point itself.The motion compensation enhances the data association link of the tracking module,improves the success rate of the association of the 3D enclosing frame,reduces the number of false associa-tions,and improves the accuracy of the multi-target tracking.The prototype validation on the relevant tracking frame-work and KITTI dataset shows that the proposed motion compensation optimization method achieves about 1%accuracy improvement.
万方数据
维普
