基于局部对比度测量的红外弱小目标恒虚警检测

发布时间：2018-06-09 16:00

本文选题：目标检测 + 弱小目标　；参考：《红外技术》2017年10期

【摘要】：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。
[Abstract]:Robust and effective small target detection algorithm is the key to the success of photoelectric tracking system. Aiming at the practical problem of detecting small infrared small targets in the air, a detection method of small infrared targets with high detection rate, low false alarm rate and short processing time is proposed based on the human visual contrast mechanism. Firstly, the original image is preprocessed by using Top-hat filter based on constant false alarm rate and adaptive threshold operation, and the suspected target area is obtained. This step can greatly reduce the computation time and maintain constant false alarm probability and predictable detection probability at the same time. Then, a novel and effective local contrast measurement operator is defined, and local saliency is calculated by introducing local self-similarity. This process can not only enhance the visual significance of the image object, but also suppress the noise. The signal-to-noise ratio (SNR) of the regional target is improved. Finally, the real target can be obtained by using a simple threshold operation based on the salient map. The qualitative and quantitative experimental results show that the proposed method has higher detection rate, lower false alarm rate and less detection time than the four existing detection algorithms. It is an effective method for detecting small and weak infrared targets in complex background.
【作者单位】：台州职业技术学院机电工程学院;西安交通大学电子与信息工程学院;
【基金】：民航联合研究基金(U13331017) 浙江省科技厅基础研究项目(016C21045) 高等学校访问学者专业发展项目(FX2012121)
【分类号】：TN219;TP391.41

【相似文献】