Research On Key Technologies Of Stable Tracking System For Airborne Optoelectronic Platform

Airborne photoelectric platform is an important payload of UAVs,which are widely used in safety monitoring,environmental monitoring and so on.In the application of the airborne optoelectronic platform stability tracking system,some problems need to be solved: The problem of the target drift when the target is tracked for a long time;The problem of the target is occluded by the cloud,the trees and so on;In haze,rain and snow and other bad weather,we can not rely on image features to track the target.This paper deals with these practical problems.In order to solve the problem of target drift due to the accumulation of errors,the mean-drift target tracking algorithm based on particle swarm optimization is adopted in the image tracker.Aiming at the problem that the target is easily occluded by clouds,trees and so on,the target tracking algorithm based on local sensitive histogram is used in the image tracker to carry out anti-occlusion tracking.The servo system and the image tracker of the airborne photoelectric platform form a closed loop,and the servo system drives the optoelectronic platform so that its axis is always aimed at the target.This paper adopts the method of geo-coordinate tracking when we can not rely on the image feature to track the target in bad weather such as fog,haze,rain and snow.The method is divided into two parts: target location and target guidance.Firstly,the latitude and longitude of the target is calculated and stored in real time according to the attitude information of the optoelectronic platform,the laser ranging value and the attitude information of the UAV.When the position of the unmanned aerial vehicle and the attitude change,the target geographic coordinate tracking system computes the azimuth and pitch angle of the photoelectric platform,and the servo system of the airborne photoelectric platform forms a closed loop with the geographic coordinate tracking system.The servo system drives the photoelectric platform,The axis is always aimed at the target.In the implementation of UAV,we need to repeatedly adjust the focal length on a number of targets for detailed observation.But when the targets are out of the field of view,we can not continue to track the targets using the image tracker.In this paper,multi-target real-time location method is used to locate multiple targets in the field of view and store the geographical coordinates of multiple targets online.After the operator has finished observing a target,the photoelectric platform can automatically track the next target by means of geographic coordinate tracking.In this paper,a multi-target positioning model and real-time correction method of zoom lens distortion are established to improve the positioning accuracy of the target.The main results of this paper are as follows:(1)The traditional Mean-shift target tracking algorithm is prone to target drift after a long time tracking.In this paper,a model of mean-drift target tracking algorithm based on domain mean particle swarm optimization is proposed.The model contains the spatial distribution information of the target color.In the process of tracking,if the background is complex,the model can be estimated by kernel density and continue to keep track the target.Compared with the traditional Mean-shift tracking algorithm,the mean-shift target tracking model based on particle swarm optimization has better tracking stability.(2)The traditional local sensitive histogram target tracking algorithm is prone to target loss.In this paper,the concept of top-level region and sub-region is proposed,and a hierarchical local sensitive histogram tracking algorithm model is established.According to the top layer template situation,we expand or reduce the template and adjust the feature update speed,effectively solve the problem of mistakenly updating target template.(3)The target positioning and geographic coordinate tracking error model are established.The factors that influence the positioning accuracy of the main target(target which is in center)are analyzed.The influence of the distortion of the zoom lens on the positioning accuracy is deduced,and the real-time distortion correction method of the zoom lens is proposed,which improves the positioning accuracy of the multi-targets.By simultaneously storing the geographical coordinates of multiple sub-targets(targets which are not in center),the photoelectric platform in turn directs each target into the field of view,effectively solving the problem that the image tracker can not keep track of multiple targets which are out of field of view.We improve the efficiency of the reconnaissance of the photoelectric platform and meet the actual needs of airborne photoelectric platform.Through the experiment,it is proved that it is feasible to use the georemal coordinate tracking method to track the targets at short focal length.