复杂背景中红外弱小目标检测技术研究

对复杂背景中红外弱小目标检测识别的一种图像处理技术进行了研究。用最大中值滤波进行预处理以抑制孤立噪声影响，采用时域方差滤波抑制背景信息，对滤波后的图像根据直方图信息进行自适应二值化处理，由多帧轨迹确定目标。用实验与基于形态学算子滤波的目标检测算法进行了比较，结果表明该算法能获得较满意的目标检测结果，两种滤波方法的组合可有效提高信噪比。

Research on Detection Technology of Small and Weak Infrared Target in Complicated Background

A kind of image treatment technology for the small and weak infrared target in complicated background was studied in this paper. The pretreatment by maximum median filtering was applied to suppress the effectiveness of the isolated noise. The variance of time-domain filtering was applied to suppress the background noise. The adoptive binarization treatment was applied to treat the filtering image according to the histogram information. The target was detected by the correlation of the multi-frame. The different target detection algorithms based on morphological filtering operator were compared by the experiment. The results showed that the target detection was satisfied using this algorithm. The integration of the two algorithms could improve the signal to noise ratio effectively.