中文摘要：

针对红外小目标视频图像信噪比低,小目标没有固定的大小形状等特性,提出了一种基于分层阈值化的红外小目标检测技术。对每一个连通域的形状、大小、边缘特性等进行评估,删除伪目标。对于一些固有的背景干扰,则利用先跟踪后检测的方法去除干扰。实验表明该算法具有良好的适应性,对于低信噪比的图片仍然具有良好的效果。

英文摘要：

In tiffs paper, infrared weak and small target detection algorithm is introduced based on layered threshold against low SNR infrared video, unfixed size and shape small target and other characteristic. Every connected domain based on its size,shape and brim charac- teristic are evaluated, then the ones which not conform to the requirements are deleted. For some inherent background interference, we can use the tracking befor detecting method to remove them. The experiment shows that this algorithm has good adaptability, and for the low SNR images still has agood effect.