基于像素驱动的特征点捕获的研究

Feature Point Caption Based on and Pixel Driving

为了实现不可见激光捕获目标物的问题,本文提出了一种基于视频图像像素驱动的激光瞄准技术。在实现将摄像机、激光测距仪和云台有机融合的基础上,将摄像机输出的像素信息转化成了云台的角度信息,并对影响转化精度的因素进行了分析,解决了摄像机镜头畸变和摄像机安装误差对像素和云台角度转换的影响。通过对各影响因素的分析,建立了特征点的运动轨迹方程。通过Matlab软件对特征点的运动轨迹分析发现,该轨迹曲线与实际运动轨迹拟合度高达99%。通过对运动轨迹方程的逆向求解,得到云台运动参数,实验结果表明,对特征点的捕获精度达到了像素级别。

In order to solve the problem of target capture by invisible laser,put forward a new laser targeting technology that based on image pixel driving.In the foundation of realizing the organic integration of camera,laser range finder and PTZ,we put pixel information of camera into the PTZ information.On the comprehensively analyzed of factors to impacting conversion precision,we solve the impacts of the conversion of PTZ angle and pixel that caused by distortion of the camera and installation error of the camera.Through the analysis of various factors,the motion trajectory equation of the feature points is established.Through analyzing the trajectory of the feature points by Matlab software,we found the fitting degree of the trajectory curve and the actual trajectory is up to 99%.Pan tilt motion parameters are obtained by the reverse solving of the trajectory equation.Experimental results show that the feature point acquisition accuracy reaches pixel level.