论文部分内容阅读
为了解决红外小目标检测问题,提出了时域红外小目标检测方法。在图像序列中背景像素、目标像素以及杂波像素的时域差分模型基础上提出了红外小目标时域检测算法,算法共分为两步:相关检测和广义似然比检测。经过相关检测后,图像序列中的噪声几乎完全被门限所抑制,只有极少数噪声像素、全部目标像素和一小部分杂波像素可以通过相关检测门限。为了进一步从通过相关检测门限的像素中检测出目标,又提出了一种新的广义似然比检测方法。这种经过改进的广义似然比检测能够进一步抑制噪声和杂波,提高检测的性能。分析表明,时域目标检测算法能够以很高的检测概率和很低的虚警概率完成目标的检测,实验方法也证实红外小目标时域检测算法具有很好的性能。
In order to solve the problem of infrared small target detection, a small time-domain infrared target detection method is proposed. Based on the time difference model of the background pixel, the target pixel and the clutter pixel in the image sequence, a small time domain infrared detection algorithm is proposed. The algorithm is divided into two steps: correlation detection and generalized likelihood ratio detection. After the relevant detection, the noise in the image sequence is almost completely suppressed by the threshold. Only a few noise pixels, all the target pixels and a small part of the clutter pixels can pass the relevant detection threshold. In order to further detect the target from the pixels that pass the relevant detection threshold, a new generalized likelihood ratio detection method is proposed. This improved generalized likelihood ratio detection can further suppress noise and clutter and improve the performance of detection. The analysis shows that the time-domain target detection algorithm can detect the target with a high detection probability and a very low false alarm probability, and the experimental method also proves that the infrared small-target time-domain detection algorithm has good performance.