红外技术, 2017, 39 (10): 940, 网络出版: 2017-12-01
基于局部对比度测量的红外弱小目标恒虚警检测
Robust Small Dim Object CFA Detection Algorithm Based on Local Contrast Measure in Aerial Complex Background
目标检测 弱小目标 恒虚警率 自适应阈值 局部对比度增强 视觉显著图 object detection infrared small object constant false alarm rate adaptive threshold local contrast enhancement visual saliency map
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
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.
姚朝霞, 谢涛. 基于局部对比度测量的红外弱小目标恒虚警检测[J]. 红外技术, 2017, 39(10): 940. YAO Zhaoxia, XIE Tao. Robust Small Dim Object CFA Detection Algorithm Based on Local Contrast Measure in Aerial Complex Background[J]. Infrared Technology, 2017, 39(10): 940.