Research On Vehicle-borne Monocular Vision Ranging Model And Its Application

The integration of monocular vision and intelligent transportation makes information acquisition based on monocular vision a hot research topic in the current transportation field.The measurement accuracy of the vehicle monocular vision ranging is not high enough and the real-time performance is not strong enough.For the problem that the influence mechanism of the error in the ranging process is not deep enough,this paper proposes a vehicle monocular ranging model based on the monocular vision principle.This article has carried out research work on the following aspects:1.A vehicle and obstacle division method based on vehicle monocular vision and the definitions of measurable area are proposed.The region of interest is extracted by the partition method in the field of view,which saves the image processing time and storage memory cost of the vehicle system.2.The original ranging model based on the principle of small hole imaging is established.The model not only includes the linear relationship between object points and pixel points,but also the direct conversion of pixel coordinates to physical coordinates.The distribution error and error generation mechanism are analyzed through experiments.Aiming at the error distribution and error generation mechanism,the improved model is proposed:the compensation scheme of the image distance influence mechanism and the compensation scheme of the optical axis tilt angle influence mechanism,and the improvement experiments are carried out according to the scheme,and the accuracy of the improved model is finally verified.3.The monocular vision ranging algorithm and flow based on the improved model are summarized,and the outdoor ranging experiment is carried out.The experimental results show that the accuracy of the algorithm is more than 99%when the static ranging is performed,and the ranging time is reduced by more than 30%than other algorithms.So the algorithm improves the accuracy and real-time performance of vehicle monocular vision ranging under static conditions.