基于加权核范数最小化的红外弱小目标检测

Infrared small and dim target detection based on weighted nuclear norm minimization

提出一种基于加权核范数最小化的红外弱小目标检测方法.该方法将原始红外图像转化为新的红外块图像模式,在红外块图像上,以鲁棒主成分分析(RPCA)为基础,将图像数据矩阵分解为一个低秩矩阵和一个稀疏矩阵;针对RPCA模型对复杂背景描述能力弱的不足,引入了加权核范数来更好地描述背景的低秩特性,并给出了相应的优化求解算法;同时,给出了一种自适应阈值分割方法,准确地从稀疏目标图像中提取出弱小目标.基于天空、海洋、山地、沙漠4种不同场景进行红外弱小目标检测,并比较了该算法和已有算法的性能,结果表明:该算法能有效地降低复杂背景边缘产生的虚警,提高目标检测准确率.

An infrared small and dim target detection method was presented based on weighted nuclear norm minimization.The original infrared image model was converted into a new infrared patch-image model.In this new infrared patch-image,the image data matrix was decomposed into a low rank matrix and a sparse matrix based on robust principal component analysis（RPCA）.To make up the poor performance of RPCA model in describing complex backgrounds,the weighted nuclear norm was introduced into RPCA for better description of the background′s low-rank property,and the corresponding optimization algorithm was also given.Besides,an adaptive threshold segmentation method was presented,which could accurately extract small and dim targets from sparse target image.The infrared small and dim target detections were carried out on four different scenes,which were sky,sea,mountains and desert.Compared with other existing methods,the performance of the proposed method can effectively lower the false alarm rate of the edge region of complex backgrounds,and raise the accuracy rate of target detection.