云杂波背景图像序列中小目标检测算法研究

云杂波背景图像序列中运动小目标的实时检测算法，是红外精确制导系统中的关键算法之一。首先对云杂波背景图像像素进行分类分析，研究了两种最大顺序滤波器和利用序列图像中像素时域剖面检测小目标的自适应方差滤波器算法，继而提出了一种适合云杂波背景的小目标检测算法。使用此算法对两组真实图像序列中运动小目标进行检测，根据对检测结果的分析，指出此算法可有效地完成检测任务，并且具有运算简单、存储量小、可并行实现、实时性好等特点。最后提出了后续工作的方向。

Detection Algorithm for Small Target in Cluttered Cloud Background Image Sequences

The real-time detection algorithm of dim small moving target in cluttered cloud background image sequences is one of key algorithms in IR precise homing systems. This paper analyzes the image pixels in cluttered cloud background image sequences, two max order filters and adaptive variance filter suited for the point target detection from temporal profile of the image sequences, and then presents an algorithm suited for the small target in the cloud image sequences. With the result of real image data experiments, it can be pointed out that it is a simple, less memory, parallelizable and real time algorithm. Finally based on the analysis of the experiment results, we point out the future work.