当前位置: X-MOL 学术IEEE Trans. Comput. Imaging › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Tensor Representation for Three-dimensional Radar Target Imaging with Sparsely Sampled Data
IEEE Transactions on Computational Imaging ( IF 4.2 ) Pub Date : 2020-01-01 , DOI: 10.1109/tci.2019.2948776
Wei Qiu , Jianxiong Zhou , Qiang Fu

Three-dimensional (3-D) radar imaging can provide additional information along elevation dimension about the target with respect to the conventional 2-D radar imaging, but usually requires a huge amount of data collected over 3-D frequency-azimuth-elevation space, which motivates us to perform 3-D imaging by using sparsely sampled data. Traditional compressive sensing (CS) based 3-D imaging methods with sparse data convert the 3-D data into a long vector, and then complete the sensing and recovery steps. This 1-D vectorized model, however, faces challenges of high computational complexity and huge memory usage and may not be viable in real applications. In this article, we solve the 3-D sparse imaging problem efficiently in a tensor way. For this aim, we firstly derive the 3-D imaging model from a tensor perspective under some assumptions. Then we review three kinds of sparse data sampling schemes that are common on the existing 3-D compressive radar imaging applications. Subsequently, with the help of prior information hidden in the radar signal, i.e., sparsity and low-rank property, we propose efficient image reconstruction algorithms for different sampling schemes to produce 3-D images with sidelobes and artifacts suppressed significantly. Finally, extensive experiments based on simulated and real-measured datasets are carried out. Results show that the proposed methods can effectively generate competitive images with small reconstruction error even when the data sampling ratio is low, which confirm the validity of proposed methods.

中文翻译:

稀疏采样数据三维雷达目标成像的张量表示

与传统的 2-D 雷达成像相比,3-D 雷达成像可以提供关于目标的沿高程维度的附加信息,但通常需要在 3-D 频率-方位-高程空间上收集大量数据,这促使我们通过使用稀疏采样数据来执行 3-D 成像。传统的基于压缩感知 (CS) 的稀疏数据 3-D 成像方法将 3-D 数据转换为长向量,然后完成感知和恢复步骤。然而,这种一维矢量化模型面临着计算复杂度高和内存使用量大的挑战,在实际应用中可能不可行。在本文中,我们以张量的方式有效地解决了 3-D 稀疏成像问题。为此,我们首先在一些假设下从张量角度推导出 3-D 成像模型。然后我们回顾了现有 3-D 压缩雷达成像应用中常见的三种稀疏数据采样方案。随后,借助隐藏在雷达信号中的先验信息,即稀疏性和低秩特性,我们针对不同的采样方案提出了有效的图像重建算法,以生成具有显着抑制旁瓣和伪影的 3-D 图像。最后,进行了基于模拟和实际测量数据集的大量实验。结果表明,即使在数据采样率较低的情况下,所提出的方法也能有效地生成重建误差较小的竞争图像,这证实了所提出方法的有效性。借助隐藏在雷达信号中的先验信息,即稀疏性和低秩特性,我们针对不同的采样方案提出了有效的图像重建算法,以产生具有显着抑制旁瓣和伪影的 3-D 图像。最后,进行了基于模拟和实际测量数据集的大量实验。结果表明,即使在数据采样率较低的情况下,所提出的方法也能有效地生成重建误差较小的竞争图像,这证实了所提出方法的有效性。借助隐藏在雷达信号中的先验信息,即稀疏性和低秩特性,我们针对不同的采样方案提出了有效的图像重建算法,以产生具有显着抑制旁瓣和伪影的 3-D 图像。最后,进行了基于模拟和实际测量数据集的大量实验。结果表明,即使在数据采样率较低的情况下,所提出的方法也能有效地生成重建误差较小的竞争图像,这证实了所提出方法的有效性。进行了基于模拟和实际测量数据集的广泛实验。结果表明,即使在数据采样率较低的情况下,所提出的方法也能有效地生成重建误差较小的竞争图像,这证实了所提出方法的有效性。进行了基于模拟和实际测量数据集的广泛实验。结果表明,即使在数据采样率较低的情况下,所提出的方法也能有效地生成重建误差较小的竞争图像,这证实了所提出方法的有效性。
更新日期:2020-01-01
down
wechat
bug