This paper studies the speckle reduction in multi-look polarimetric synthetic aperture radar (SAR) image. A multi-look polarimetric whitening filtering (MPWF) method is presented and extended to form a fully polarimet...This paper studies the speckle reduction in multi-look polarimetric synthetic aperture radar (SAR) image. A multi-look polarimetric whitening filtering (MPWF) method is presented and extended to form a fully polarimetric filter with multi-channel output. The paper also quantifies the speckle reduction amount achievable by the MPWF, and compares the MPWF with the span, weighting and power equalization methods. Experimental results with the NASA/JPL L-band 4-look polarimetric SAR data verify the effectiveness and superiority of the MPWF, and show that the MPWF is of great advantage for enhancing SAR image classification.展开更多
In this paper, a new maximum likelihood (ML) classification algorithm is proposed to classify the multi-look polarimetric synthetic aperture radar (SAR) imagery. Experimental results with the NASA/JPL airborne L-band ...In this paper, a new maximum likelihood (ML) classification algorithm is proposed to classify the multi-look polarimetric synthetic aperture radar (SAR) imagery. Experimental results with the NASA/JPL airborne L-band polarimetric SAR data demonstrate the effectiveness of the new algorithm. Furthermore, when using the algorithm in the classifications with subsets of the multi-look polarimetric SAR data, the polarization-channel optimization for the terrain type classification is implemented.展开更多
With a multiplicative speckle model, this paper shows the multi-look polarimetric synthetic aperture radar (SAR) data obeys a generalized K-distribution. To validate this distribution model, the multi-look intensity K...With a multiplicative speckle model, this paper shows the multi-look polarimetric synthetic aperture radar (SAR) data obeys a generalized K-distribution. To validate this distribution model, the multi-look intensity K-distribution is particularly tested. The relationship between the heterogeneity coefficient of the scene and the proper statistical model is experimentally established. In addition, based on the results of the statistical analysis, an adaptive classification scheme is presented, and the improved classification shows the importance of the statistical analysis.展开更多
为了解决施工场景下安全帽佩戴检测时,由于人员密集、遮挡和复杂背景等原因造成的小目标漏检和错检的问题,提出一种基于YOLOv8n的双重注意力机制的跨层多尺度安全帽佩戴检测算法。首先,设计微小目标检测头,以提高模型对小目标的检测能力...为了解决施工场景下安全帽佩戴检测时,由于人员密集、遮挡和复杂背景等原因造成的小目标漏检和错检的问题,提出一种基于YOLOv8n的双重注意力机制的跨层多尺度安全帽佩戴检测算法。首先,设计微小目标检测头,以提高模型对小目标的检测能力;其次,在特征提取网络中嵌入双重注意力机制,从而更加关注复杂场景下目标信息的特征捕获;然后,将特征融合网络替换成重参数化泛化特征金字塔网络(RepGFPN)改进后的跨层多尺度特征融合结构S-GFPN(Selective layer Generalized Feature Pyramid Network),以实现小目标特征层信息和其他特征层的多尺度融合,并建立长期的依赖关系,从而抑制背景信息的干扰;最后,采用MPDIOU(Intersection Over Union with Minimum Point Distance)损失函数来解决尺度变化不敏感的问题。在公开数据集GDUT-HWD上的实验结果表明,改进后的模型比YOLOv8n的mAP@0.5提升了3.4个百分点,对蓝色、黄色、白色和红色安全帽的检测精度分别提升了2.0、1.1、4.6和9.1个百分点,在密集、遮挡、小目标、反光和黑暗这5类复杂场景下的可视化检测效果也优于YOLOv8n,为实际施工场景中安全帽佩戴检测提供了一种有效方法。展开更多
针对交通路口图像复杂,小目标难测且目标之间易遮挡以及天气和光照变化引发的颜色失真、噪声和模糊等问题,提出一种基于YOLOv9(You Only Look Once version 9)的交通路口图像的多目标检测算法ITD-YOLOv9(Intersection Target Detection-...针对交通路口图像复杂,小目标难测且目标之间易遮挡以及天气和光照变化引发的颜色失真、噪声和模糊等问题,提出一种基于YOLOv9(You Only Look Once version 9)的交通路口图像的多目标检测算法ITD-YOLOv9(Intersection Target Detection-YOLOv9)。首先,设计CoT-CAFRNet(Chain-of-Thought prompted Content-Aware Feature Reassembly Network)图像增强网络,以提升图像质量,并优化输入特征;其次,加入通道自适应特征融合(iCAFF)模块,以增强小目标及重叠遮挡目标的提取能力;再次,提出特征融合金字塔结构BiHS-FPN(Bi-directional High-level Screening Feature Pyramid Network),以增强多尺度特征的融合能力;最后,设计IF-MPDIoU(Inner-Focaler-Minimum Point Distance based Intersection over Union)损失函数,以通过调整变量因子,聚焦关键样本,并增强泛化能力。实验结果表明,在自制数据集和SODA10M数据集上,ITD-YOLOv9算法的检测精度分别为83.8%和56.3%,检测帧率分别为64.8 frame/s和57.4 frame/s。与YOLOv9算法相比,ITD-YOLOv9算法的检测精度分别提升了3.9和2.7个百分点。可见,所提算法有效实现了交通路口的多目标检测。展开更多
文摘This paper studies the speckle reduction in multi-look polarimetric synthetic aperture radar (SAR) image. A multi-look polarimetric whitening filtering (MPWF) method is presented and extended to form a fully polarimetric filter with multi-channel output. The paper also quantifies the speckle reduction amount achievable by the MPWF, and compares the MPWF with the span, weighting and power equalization methods. Experimental results with the NASA/JPL L-band 4-look polarimetric SAR data verify the effectiveness and superiority of the MPWF, and show that the MPWF is of great advantage for enhancing SAR image classification.
基金This work was performed at Alenia Spazio,Rome,Italy.It was a part of the cooperation project between Alenia Spazio and University of Electronic Science and Technology of China,Chengdu,China
文摘In this paper, a new maximum likelihood (ML) classification algorithm is proposed to classify the multi-look polarimetric synthetic aperture radar (SAR) imagery. Experimental results with the NASA/JPL airborne L-band polarimetric SAR data demonstrate the effectiveness of the new algorithm. Furthermore, when using the algorithm in the classifications with subsets of the multi-look polarimetric SAR data, the polarization-channel optimization for the terrain type classification is implemented.
基金This work was performed at Alenia Spazio, Rome, Italy. It was a part of the cooperation project between Alenia Spazio and University of Electronic Science and Technology of China,Chengdu, China
文摘With a multiplicative speckle model, this paper shows the multi-look polarimetric synthetic aperture radar (SAR) data obeys a generalized K-distribution. To validate this distribution model, the multi-look intensity K-distribution is particularly tested. The relationship between the heterogeneity coefficient of the scene and the proper statistical model is experimentally established. In addition, based on the results of the statistical analysis, an adaptive classification scheme is presented, and the improved classification shows the importance of the statistical analysis.
文摘为了解决施工场景下安全帽佩戴检测时,由于人员密集、遮挡和复杂背景等原因造成的小目标漏检和错检的问题,提出一种基于YOLOv8n的双重注意力机制的跨层多尺度安全帽佩戴检测算法。首先,设计微小目标检测头,以提高模型对小目标的检测能力;其次,在特征提取网络中嵌入双重注意力机制,从而更加关注复杂场景下目标信息的特征捕获;然后,将特征融合网络替换成重参数化泛化特征金字塔网络(RepGFPN)改进后的跨层多尺度特征融合结构S-GFPN(Selective layer Generalized Feature Pyramid Network),以实现小目标特征层信息和其他特征层的多尺度融合,并建立长期的依赖关系,从而抑制背景信息的干扰;最后,采用MPDIOU(Intersection Over Union with Minimum Point Distance)损失函数来解决尺度变化不敏感的问题。在公开数据集GDUT-HWD上的实验结果表明,改进后的模型比YOLOv8n的mAP@0.5提升了3.4个百分点,对蓝色、黄色、白色和红色安全帽的检测精度分别提升了2.0、1.1、4.6和9.1个百分点,在密集、遮挡、小目标、反光和黑暗这5类复杂场景下的可视化检测效果也优于YOLOv8n,为实际施工场景中安全帽佩戴检测提供了一种有效方法。
文摘针对交通路口图像复杂,小目标难测且目标之间易遮挡以及天气和光照变化引发的颜色失真、噪声和模糊等问题,提出一种基于YOLOv9(You Only Look Once version 9)的交通路口图像的多目标检测算法ITD-YOLOv9(Intersection Target Detection-YOLOv9)。首先,设计CoT-CAFRNet(Chain-of-Thought prompted Content-Aware Feature Reassembly Network)图像增强网络,以提升图像质量,并优化输入特征;其次,加入通道自适应特征融合(iCAFF)模块,以增强小目标及重叠遮挡目标的提取能力;再次,提出特征融合金字塔结构BiHS-FPN(Bi-directional High-level Screening Feature Pyramid Network),以增强多尺度特征的融合能力;最后,设计IF-MPDIoU(Inner-Focaler-Minimum Point Distance based Intersection over Union)损失函数,以通过调整变量因子,聚焦关键样本,并增强泛化能力。实验结果表明,在自制数据集和SODA10M数据集上,ITD-YOLOv9算法的检测精度分别为83.8%和56.3%,检测帧率分别为64.8 frame/s和57.4 frame/s。与YOLOv9算法相比,ITD-YOLOv9算法的检测精度分别提升了3.9和2.7个百分点。可见,所提算法有效实现了交通路口的多目标检测。
