| Compared with active ranging,vision-based passive ranging has the characteristics of rich information,non-destructive,non-contact,and high concealment,which makes it used in more and more fields.Among them,the monocular passive ranging uses a single camera for ranging,which makes it has the advantages of far measurable distance,simple structure and low price.However,in the case of a long distance to be measured,the target is often a point.The monocular passive ranging has an ill-posed problem.Regarding the issue above,this paper uses the method of trajectory and line of sight for ranging.The main research contents are as follows:(1)This paper studies the basic theories of visual measurement and analyzes the observability issues of monocular passive ranging technology.The more commonly used general inferences in engineering practice are obtained,which serve as the basis for the ranging algorithm models and accuracy analysis.(2)Aiming at the ill-posed problem of monocular point target ranging,the uniform speed trajectory constraint method in 3D space and the trajectory smooth constraint method based on unit sight vector constraints are proposed.The ill-posed problem is solved by parameterizing the trajectory of the target,or by minimizing the trajectory smoothing term.After simulation verification,the algorithms in this paper effectively solve the problems of ranging for the target of uniform motion and complex trajectory,and show good anti-interference.(3)Aiming at the ranging accuracy of the monocular ranging,the transmission law of lineof-sight angle error have been studied.Aiming at the characteristics of two algorithms,the simulation analysis of the factors affecting the accuracy is carried out.A calculation method of accuracy based on statistical analysis is proposed.According to the individual error distribution characteristics,a large sample of random data is established,and the angle measurement and ranging accuracy are calculated.The theoretical ranging accuracy of algorithms in this paper is±(L×1.45%)and ±(L×1.81%)for targets at 3～4km after simulation verification,which achieves the performance index of ranging accuracy better than 2%.According to the final simulation results,the ranging algorithms in this paper have been verified to be effective for ranging moving point targets and providing the theoretical ranging accuracy index. |
