基于区域生长与区域覆盖比的红外弱小目标检测

Infrared small target detection based on region growth and region coverage ratio

为了提高复杂背景下红外弱小目标的检测能力,提出了一种基于区域生长与区域覆盖比的新方法。本文中主要针对目标的三个特征:1)目标的最大灰度值与邻域背景之间存在较为明显的灰度间隙;2)目标在局部范围内具有较高的灰度值;3)目标像素分布较为紧凑,建立了一种新的小目标检测算法。首先,对原始图像计算局部对比度并筛选出候选种子点。其次,对每个候选种子点在原始图像上进行阈值区域生长算法并计算得到区域覆盖比(RCR)。然后,使用自适应尺寸的三层窗口计算得到自适应灰度差(AGVD)。最后,采用阈值分割方法分离出真实目标。通过在真实测试数据集上的实验表明,与现有算法相比,所提出的算法具有较高的检测精度和较低的虚警率。

To improve the detection ability of infrared small targets in complex backgrounds,this paper proposes a novel method at the region growth and region coverage ratio.This paper focuses mainly on the following three characteristics of the target:1)there being an obvious gray gap between the maximum gray value of the target and the neighborhood background;2)the targets having high gray values in the local range;3)the target pixel distribution being compact.A new small target detection algorithm is established in this paper.Firstly,the local contrast of the original image and screen out the candidate seed points is calculate.Secondly,the threshold region growth algorithm is performed on the original image for each candidate seed point,with the region coverage ratio(RCR)calculated.Then,the adaptive gray value difference(AGVD)is calculated using the three-layer window in an adaptive size.Finally,the threshold segmentation method is used to obtain the real targets.Experiments on real test data sets show that the proposed algorithm has a higher detection accuracy and a lower false alarm rate than the existing algorithms.