Distance Estimation of Monocular Based on Vehicle Pose Information

Abstract

In order to measure the distance between our vehicle and the target vehicle by monocular vision and eliminate the estimation error bring by changing of vehicle pose, we propose the distance estimation method based on the vehicle pose information, which can be used to eliminate the error of distance estimation effectively cause by the change of the pitch angle and roll angle of unmanned vehicle. In addition, the pose information could also help us estimate whether the vehicle is in a slope, thus engage in distance estimation for the vehicles. Several groups of data was collected for the experiments. And the result prove the validity of the algorithm in distance estimation.