A precise method for accurately tracking dim- small targets, based on spectral fingerprint is proposed where traditional full color tracking seems impossible. A fingerprint model is presented to adequately extract spe...A precise method for accurately tracking dim- small targets, based on spectral fingerprint is proposed where traditional full color tracking seems impossible. A fingerprint model is presented to adequately extract spectral features. By creating a multidimensional feature space and extending the limited RGB information to the hyperspectral information, the improved precise tracking model based on a nonparamet- ric kernel density estimator is built using the probability his- togram of spectral features. A layered particle filter algorithm for spectral tracking is presented to avoid the object jumping abruptly. Finally, experiments are conducted that show that the tracking algorithm with spectral fingerprint features is ac- curate, fast, and robust. It meets the needs of dim-small target tracking adequately.展开更多
针对单帧复杂背景红外图像点目标检测算法存在复杂背景下处理效果不理想、处理时间长的问题,提出了一种层次卷积滤波检测算法。主要分为两个部分:第一,根据红外小目标特性,设计一种层次卷积滤波的算子,对图像进行滤波处理,实现图像中小...针对单帧复杂背景红外图像点目标检测算法存在复杂背景下处理效果不理想、处理时间长的问题,提出了一种层次卷积滤波检测算法。主要分为两个部分:第一,根据红外小目标特性,设计一种层次卷积滤波的算子,对图像进行滤波处理,实现图像中小目标的增效和背景抑制的效果;第二,采用基于最大值的自适应阈值方法,对图像进行二值化操作,过滤背景杂波,最终提取到待检测的目标。在大量不同背景红外图像中进行实验,论文算法在背景抑制因子和信噪比增益的性能量化结果上优于现有5种典型红外弱小目标检测算法的性能结果,且平均处理时间仅为高斯拉普拉斯(Laplacian of Gaussian,LoG)滤波算法的30.42%。通过实验对比,表明该层次卷积滤波算法可以有效解决在不同复杂背景下的红外图像中对小目标检测的问题。展开更多
红外弱小目标检测技术是红外探测系统的核心技术之一。针对远距离复杂场景下红外弱小目标对比度低、信噪比低和纹理特征稀疏分散导致目标检测率低的问题,提出一种融合注意力机制和改进YOLOv3的红外弱小目标检测算法。首先,在YOLOv3的基...红外弱小目标检测技术是红外探测系统的核心技术之一。针对远距离复杂场景下红外弱小目标对比度低、信噪比低和纹理特征稀疏分散导致目标检测率低的问题,提出一种融合注意力机制和改进YOLOv3的红外弱小目标检测算法。首先,在YOLOv3的基础上,用更大尺度的检测头替换最小尺度的检测头,在保证推理速度的基础上有效提升了红外图像中小目标的检测概率。然后,在检测头之前设计了Infrared Attention模块,通过通道间的信息交互,抽取出更加关键重要的信息供网络学习。最后,用完全交并比损失(Complete IoU Loss)替代交并比损失(Intersection over Union Loss)来衡量预测框的检测能力,通过梯度回传实现更好的模型训练。实验结果表明,提出的YOLOv3-DCA能完成多种场景下红外弱小目标的检测任务,且检测准确率、召回率、F1和平均准确率分别达到91.8%、88.8%、93.0%和88.8%,平均准确率比YOLOv3基线提升约7%,与主流的SSD、CenterNet和YOLOv4模型对比平均准确率也取得了目前最优。展开更多
文摘A precise method for accurately tracking dim- small targets, based on spectral fingerprint is proposed where traditional full color tracking seems impossible. A fingerprint model is presented to adequately extract spectral features. By creating a multidimensional feature space and extending the limited RGB information to the hyperspectral information, the improved precise tracking model based on a nonparamet- ric kernel density estimator is built using the probability his- togram of spectral features. A layered particle filter algorithm for spectral tracking is presented to avoid the object jumping abruptly. Finally, experiments are conducted that show that the tracking algorithm with spectral fingerprint features is ac- curate, fast, and robust. It meets the needs of dim-small target tracking adequately.
文摘针对单帧复杂背景红外图像点目标检测算法存在复杂背景下处理效果不理想、处理时间长的问题,提出了一种层次卷积滤波检测算法。主要分为两个部分:第一,根据红外小目标特性,设计一种层次卷积滤波的算子,对图像进行滤波处理,实现图像中小目标的增效和背景抑制的效果;第二,采用基于最大值的自适应阈值方法,对图像进行二值化操作,过滤背景杂波,最终提取到待检测的目标。在大量不同背景红外图像中进行实验,论文算法在背景抑制因子和信噪比增益的性能量化结果上优于现有5种典型红外弱小目标检测算法的性能结果,且平均处理时间仅为高斯拉普拉斯(Laplacian of Gaussian,LoG)滤波算法的30.42%。通过实验对比,表明该层次卷积滤波算法可以有效解决在不同复杂背景下的红外图像中对小目标检测的问题。
文摘红外弱小目标检测技术是红外探测系统的核心技术之一。针对远距离复杂场景下红外弱小目标对比度低、信噪比低和纹理特征稀疏分散导致目标检测率低的问题,提出一种融合注意力机制和改进YOLOv3的红外弱小目标检测算法。首先,在YOLOv3的基础上,用更大尺度的检测头替换最小尺度的检测头,在保证推理速度的基础上有效提升了红外图像中小目标的检测概率。然后,在检测头之前设计了Infrared Attention模块,通过通道间的信息交互,抽取出更加关键重要的信息供网络学习。最后,用完全交并比损失(Complete IoU Loss)替代交并比损失(Intersection over Union Loss)来衡量预测框的检测能力,通过梯度回传实现更好的模型训练。实验结果表明,提出的YOLOv3-DCA能完成多种场景下红外弱小目标的检测任务,且检测准确率、召回率、F1和平均准确率分别达到91.8%、88.8%、93.0%和88.8%,平均准确率比YOLOv3基线提升约7%,与主流的SSD、CenterNet和YOLOv4模型对比平均准确率也取得了目前最优。