This paper proposes an application of compressive imaging systems to the problem of wide-area video surveillance systems. A parallel coded aperture compressive imaging system and a corresponding motion target detectio...This paper proposes an application of compressive imaging systems to the problem of wide-area video surveillance systems. A parallel coded aperture compressive imaging system and a corresponding motion target detection algorithm in video using compressive image data are developed. Coded masks with random Gaussian, Toeplitz and random binary are utilized to simulate the compressive image respectively. For compressive images, a mixture of the Gaussian distribution is applied to the compressed image field to model the background. A simple threshold test in compressive sampling image is used to declare motion objects. Foreground image retrieval from underdetermined measurement using the total variance optimization algorithm is explored. The signal-to-noise ratio （SNR） is employed to evaluate the image quality recovered from the compressive sampling signals, and receiver operation characteristic （ROC） curves are used to quantify the performance of the motion detection algorithm. Experimental results demonstrate that the low dimensional compressed imaging representation is sufficient to determine spatial motion targets. Compared with the random Gaussian and Toeplitz mask, motion detection algorithms using the random binary phase mask can yield better detection results. However using the random Gaussian and Toeplitz phase mask can achieve high resolution reconstructed images.展开更多
The encoding aperture snapshot spectral imaging system,based on the compressive sensing theory,can be regarded as an encoder,which can efficiently obtain compressed two-dimensional spectral data and then decode it int...The encoding aperture snapshot spectral imaging system,based on the compressive sensing theory,can be regarded as an encoder,which can efficiently obtain compressed two-dimensional spectral data and then decode it into three-dimensional spectral data through deep neural networks.However,training the deep neural net⁃works requires a large amount of clean data that is difficult to obtain.To address the problem of insufficient training data for deep neural networks,a self-supervised hyperspectral denoising neural network based on neighbor⁃hood sampling is proposed.This network is integrated into a deep plug-and-play framework to achieve self-supervised spectral reconstruction.The study also examines the impact of different noise degradation models on the fi⁃nal reconstruction quality.Experimental results demonstrate that the self-supervised learning method enhances the average peak signal-to-noise ratio by 1.18 dB and improves the structural similarity by 0.009 compared with the supervised learning method.Additionally,it achieves better visual reconstruction results.展开更多
合成孔径雷达(Synthetic aperture radar,SAR)图像因为相干斑现象和目标响应的空间变化呈现出一种纹理特性,局部二进编码等局部图像特征在光学纹理描述中获得较好的结果,但光学纹理特征在描述SAR图像纹理特性中因为相干成像特性往往失效...合成孔径雷达(Synthetic aperture radar,SAR)图像因为相干斑现象和目标响应的空间变化呈现出一种纹理特性,局部二进编码等局部图像特征在光学纹理描述中获得较好的结果,但光学纹理特征在描述SAR图像纹理特性中因为相干成像特性往往失效.本文在前期工作纹理特征框架的基础上,提出了一种局部重要性采样二进编码的SAR图像纹理特征(Feature extraction based on local important sampling binary,LISBF)描述方法:首先,利用样本图像对局部采样位置进行随机自适应采样,基于重要性采样(Important sample,IS)方法输出递归学习位置结果;然后,利用学习出的纹理重要采样点对进行二进特征编码;最后,通过映射和统计生成描述算子.该特征较固定位置采样能够获取更大范围信息,同时能通过采样避免特征维数的急剧增大;通过自适应学习重要性关键点较随机采样更容易捕捉纹理固有信息;较好地适应了SAR图像极低信噪比和斑点现象的纹理.本文将该特征用于真实图像和标准纹理库的分类研究,实验结果证明了该特征的有效性.展开更多
基金supported by the National Natural Science Foundation of China (61271375)BIT Foundation (2012CX04054)
文摘This paper proposes an application of compressive imaging systems to the problem of wide-area video surveillance systems. A parallel coded aperture compressive imaging system and a corresponding motion target detection algorithm in video using compressive image data are developed. Coded masks with random Gaussian, Toeplitz and random binary are utilized to simulate the compressive image respectively. For compressive images, a mixture of the Gaussian distribution is applied to the compressed image field to model the background. A simple threshold test in compressive sampling image is used to declare motion objects. Foreground image retrieval from underdetermined measurement using the total variance optimization algorithm is explored. The signal-to-noise ratio （SNR） is employed to evaluate the image quality recovered from the compressive sampling signals, and receiver operation characteristic （ROC） curves are used to quantify the performance of the motion detection algorithm. Experimental results demonstrate that the low dimensional compressed imaging representation is sufficient to determine spatial motion targets. Compared with the random Gaussian and Toeplitz mask, motion detection algorithms using the random binary phase mask can yield better detection results. However using the random Gaussian and Toeplitz phase mask can achieve high resolution reconstructed images.
基金Supported by the Zhejiang Provincial"Jianbing"and"Lingyan"R&D Programs(2023C03012,2024C01126)。
文摘The encoding aperture snapshot spectral imaging system,based on the compressive sensing theory,can be regarded as an encoder,which can efficiently obtain compressed two-dimensional spectral data and then decode it into three-dimensional spectral data through deep neural networks.However,training the deep neural net⁃works requires a large amount of clean data that is difficult to obtain.To address the problem of insufficient training data for deep neural networks,a self-supervised hyperspectral denoising neural network based on neighbor⁃hood sampling is proposed.This network is integrated into a deep plug-and-play framework to achieve self-supervised spectral reconstruction.The study also examines the impact of different noise degradation models on the fi⁃nal reconstruction quality.Experimental results demonstrate that the self-supervised learning method enhances the average peak signal-to-noise ratio by 1.18 dB and improves the structural similarity by 0.009 compared with the supervised learning method.Additionally,it achieves better visual reconstruction results.
文摘合成孔径雷达(Synthetic aperture radar,SAR)图像因为相干斑现象和目标响应的空间变化呈现出一种纹理特性,局部二进编码等局部图像特征在光学纹理描述中获得较好的结果,但光学纹理特征在描述SAR图像纹理特性中因为相干成像特性往往失效.本文在前期工作纹理特征框架的基础上,提出了一种局部重要性采样二进编码的SAR图像纹理特征(Feature extraction based on local important sampling binary,LISBF)描述方法:首先,利用样本图像对局部采样位置进行随机自适应采样,基于重要性采样(Important sample,IS)方法输出递归学习位置结果;然后,利用学习出的纹理重要采样点对进行二进特征编码;最后,通过映射和统计生成描述算子.该特征较固定位置采样能够获取更大范围信息,同时能通过采样避免特征维数的急剧增大;通过自适应学习重要性关键点较随机采样更容易捕捉纹理固有信息;较好地适应了SAR图像极低信噪比和斑点现象的纹理.本文将该特征用于真实图像和标准纹理库的分类研究,实验结果证明了该特征的有效性.
