基于线检测器的红外弱小目标检测

Dim and Small IR Targets Detection Based on Linear Detector

复杂背景下低信噪比弱小目标的检测仍然是当今的一个难题,尤其是在复杂背景的抑制方面,针对红外小目标的特点,以及红外小目标实时性检测的要求,提出了一种检测低信噪比的复杂背景情况下的红外图像点目标的新的算法。首先,用一个线检测器,对红外图像进行处理,经过其处理后的图像,不仅可以大幅度抑制背景,而且还能增强目标,极大程度的提高了图像的信噪比;其次,对处理后的图像用二维的Otsu自动阈值分割法进行分割,检测出点目标。通过对红外图像的实验,证明该算法的有效性。

It is a difficult problem to detect weak and small targets in complicated background, especially in the field of suppressing the complicated background. In view of the behavior of the small targets and the requirement of real - time detection ,this paper presents a new algorithm for detecting point targets in low SNR and complicated back- ground IR image. Firstly,preprocessing the IR image with a linear detector,the preprocessed images not only can sup- press background but also can enhance the targets, so the SNR is also improved. Secondly, applying two - dimensional Otsu threshold algorithm to segment the preprocessed IR images, thus detecting small IR targets clearly. Experiments demonstrate that this method can detect small IR targets accurately.