基于连续帧的红外小目标检测技术

摘要

An infrared small target detecting algorithm is presented with consecutive frames. The algorithm uses the enhanced simplified high-pass filter to the initial image, and makes binarization with the iteration of histogram. Finally, the aimed target is confirmed after consecutive frames. Simulations show that the algorithm can realize the fast and stable detection to the infrared small target with low SNR and complex background.%提出了一种基于连续帧的远距红外小目标检测技术.通过对原始红外图像进行增强型简化高通滤波提高图像信噪比,用基于直方图迭代得到的阈值二值化后,得到小目标图像.为了确保目标检测的可靠性,利用连续多帧确定截获的目标.仿真结果表明该算法可实现在低信噪比复杂背景中对远距小目标的快速稳定截获.