摘要
针对靶场飞行试验光电经纬仪测量图像存在海背景复杂、跟踪不稳定、粗差噪声多等问题,常规图像目标检测跟踪算法计算量大,虚警率高,判读效率低。文中充分利用经纬仪测量图像目标测角信息的相关性和趋势性,提出一种基于抗差最小二乘的测量图像目标位置快速预测方法,通过估计经纬仪测角信息达到预测图像目标位置的目的。经试验验证:该方法模型简单,计算速度快,预测误差在2个像素以内,计算量较传统方法至少降低2~3倍,能够大幅提高靶场测量图像判读效率,具有推广应用价值。
To solve the problems of complex sea background,unstable tracking,and many gross errors and noises in the photoelectric theodolite measurement image of the shooting range flight test,the conventional image target detection and tracking algorithm has a large amount of calculation,high false alarm rate,and low interpretation efficiency.This paper makes full use of the correlation and trend of the angle measurement information of the theodolite measurement image target,and proposes a fast prediction method of the target position of the measurement image based on the robust least squares.The goal of predicting the target position of the image is achieved by estimating the angle measurement information of the theodolite.After experimental verification,this method has simple model,fast calculation speed,prediction error within 2 pixels,and computational complexity could be reduced by at least 2-3 times compared to traditional methods,which can significantly improve the efficiency of target range measurement image interpretation and has promotion and application value.
作者
韩先平
HAN Xian-ping(Unit 92941 of the PLA,Huludao 125000,Liaoning Province,China)
出处
《信息技术》
2025年第9期52-57,共6页
Information Technology
基金
海军装备建设科研项目(HJ-301-2019-1501)。
关键词
抗差估计
最小二乘
等价权函数
粗差
测角信息
robust estimation
least squares
equivalent weight function
gross error
angle measurement information
