基于隔帧差分向量无穷范数的运动弱小目标的检测

Dim Moving Target Detection Based on Infinite Norm of the Discontinuous Framed Difference Vector

提出了运用隔帧差分向量无穷范数检测红外图像序列中运动弱小目标一种新算法。算法以隔帧差分为基础,该处理系统不仅能够探测到帧间位移不小于1个像元的点目标,而且可以探测到帧间位移小于1个像元而多帧累积位移大于1个像元的运动点目标,使算法探测与识别目标的能力大大提高,为不同速度多目标的检测提供了新的可能。仿真实验结果表明,该算法具有较高的检测率和良好的实时特性,能有效地检测出低信比红外图像序列中的弱小运动目标。

A new algorithm for detecting dim moving target in IR image sequence by the infinite norm of discontinuous frame difference vector is proposed in this paper. Based on discontinuous frame difference, the detection system can detect not only the dim moving target whose displacement is no less than one pixel between two continuous frames, but also the dim moving one whose displacement is no more than one pixel between two continuous frames and whose gathered displacement of multiple continuous frames is larger than one pixel. These provide a new solution to detect multiple targets with different velocities in IR image sequence. Simulations of real IR image sequence with low SNR prove the algorithm can effectively detect dim moving target, and has high detection probability and excellent real time performance.