基于双重增强局部对比度的红外小目标检测

Detection of infrared small targets based on dual-enhanced local contrast measure

红外探测具有穿透力强、探测距离远和隐蔽性好等特点,在红外小目标检测领域应用广泛。本文基于人类视觉机制的局部对比度方法,提出一种基于双重增强局部对比度的红外小目标检测方法,先以局部灰度均值对比度对噪声和背景进行抑制,再由相对灰度梯度和局部方差对比结合对目标进行增强,最终采用阈值分割的方式实现红外小目标的检测。算法对不同天空场景序列目标均有较高的检测率和较低的虚警率,前者最高可达100%,后者为0;且相同复杂场景的目标检测性能相比其他同类方法明显提高,信杂比增益与背景抑制因子也显著提升,目标检测性能优异。

Infrared detection is widely used in the field of infrared small target detection due to its strong penetration,long detection distance and good concealment.In this paper,an infrared small target detection method based on dualenhanced local contrast measurement is proposed on the basis of the local contrast method of human visual mechanism.Firstly,the noise and background are suppressed by local gray mean contrast,followed by target enhancement by a combination of relative grey scale gradient and local variance contrast,and the infrared small target is detected by threshold segmentation finally.The algorithm has higher detection rate and lower false alarm rate for different sky scene sequences,the former is up to 100%,and the latter is 0.The target detection performance of the same complex scene is significantly improved compared with other similar methods,signal-to-clutter ratio gain and background suppression factor are also significantly improved,resulting in excellent target detection performance.