鲁棒有效的弱小目标检测算法是光电跟踪系统成功的关键。本文针对空中远距离红外弱小目标检测的实际问题，在人类视觉对比机制基础上提出了一种检测率高、误报率低、处理时间短的红外小目标检测方法。首先，利用基于恒虚警率的Top-hat 滤波和自适应阈值操作对原始图像进行预处理，得到疑似目标区域，该步骤可大大减少计算时间，同时保持恒定的虚警概率和可预测的检测概率；然后，定义了一种新颖有效的局部对比度测量算子，并引入图像局部的自相似性计算局部显著图，该过程不仅可以增强图像目标的视觉显著性，同时还可以抑制噪声，提高区域目标的信噪比；最后，在显著图基础上，利用简单的阈值操作就可以获得真实目标。定性定量实验结果表明，本文提出的方法与4 种现有检测算法相比，具有更高的检测率、更低的虚警率和更少的检测时间，是复杂背景下红外弱小目标检测的有效方法。

A robust and effective small dim object detection algorithm is the key to the success of an infrared tracking system. To help solve practical tracking problems, we propose a small dim infrared object detection algorithm with a high detection rate, a low false alarm rate, and a short processing time. First, using Top-hat filter and adaptive threshold operation based on a constant false alarm rate, original images are pre-processed to obtain the suspected object area, greatly reducing computation time and detection probability, while maintaining a constant false alarm probability. Second, we define a novel and effective local contrast measurement operator, and introduce a local self-similarity measure of a local saliency map, enhancing not only visual saliency, but also improving signal-to-noise ratio. Finally, a simple threshold operation in the saliency map can be used to detect the real object. Many simulation results show that our proposed algorithm is superior to existing detection algorithms in a complex background.