摘要： 提出一种基于卡尔曼滤波的弱小目标实时检测与跟踪方法。计算每帧图像上所有星点到参考星点的距离，利用目标与背景恒星运动特性上的差异检测出运动目标。针对漏检问题，采用卡尔曼滤波算法估计目标在漏检帧上的位置，通过对图像的重分割寻找丢失目标，利用目标的运动信息建立连续的目标链。实验结果表明，该方法能实现高检测率和低虚警率的实时检测。

关键词: 空间图像, 弱小目标, 星点距, 卡尔曼滤波, 实时检测, 目标匹配

Abstract: This paper presents a real-time detection and tracking method for dim-small target based on Kalman Filtering(KF). In adjoining frames same reference stars are selected and it calculates the distance of every star to the reference stars. Because the star points in the background have different movement from the targets, the true targets can be found from the stars. To solve the problem of targets loss, KF is used to forecast the position and the picture is segmented to find the lost target, and target chains are built with the movement stability. Experimental results show that the method can perfectly meet the requirements of the real-time space target detection with a high detection probability and a low false alarm rate.

Key words: space image, dim-small target, star distance, Kalman Filtering(KF), real-time detection target matching

中图分类号: 

• TP391.4