水下探测中基于压缩感知的合成孔径成像
发布时间：2018-11-18 09:15
【摘要】：随着我国海洋事业的蓬勃发展,对水下成像质量要求日渐增高,具备高分辨率成像能力的合成孔径声纳设备逐渐成为研究热点。为了获得高分辨率成像,通常需要采集大量的水下数据,会给相应的硬件设备及发送测量数据的通信系统带来极大的挑战。因此,研究如何在保证成像质量的同时降低所需数据量具有重要的现实意义。 压缩感知理论(Compressed Sensing)利用信号的稀疏特性,可以将无失真重建信号的最小采样频率降低到奈奎斯特频率以下,可用于解决成像数据量大的问题。在本文中,主要探讨研究了两种应用压缩感知的合成孔径成像算法。由于声纳信号的稀疏矩阵位于复数域,本文给出了重构算法复数域的实现形式,最后利用点目标的成像仿真实验验证了算法性能。 本文的主要工作如下： (1)详细介绍了压缩感知基本理论和基础知识,主要包括信号的稀疏、非线性测量及原始信号的重建三个部分。研究合成孔径声纳成像原理及经典的距离多普勒成像算法,并完成了四点目标的成像仿真； (2)为解决距离向较大数据量的问题,研究了具有保相性的距离向压缩感知的成像算法。实验结果显示,该算法在距离向仅用20%的原始数据,即可获得良好的成像效果;算法本身可以起到抑制旁瓣的作用,且具有良好的抗噪性能。为了在不产生方位模糊的前提下获得更远的观察距离,研究了二维降采样压缩感知成像算法,该算法在保证成像质量的同时,可进一步降低数据量； (3)探讨两种压缩感知成像算法中重建算法的复数域实现形式,即经典算法的分解式复数域重构形式及可在复数域直接求解的SPGL1算法,结果表明,SPGL1算法不仅可以节省存储空间,而且抗噪性能较好。
[Abstract]:With the rapid development of marine industry in China, the requirement of underwater imaging quality is increasing, and synthetic aperture sonar equipment with high resolution imaging capability has gradually become a research hotspot. In order to obtain high resolution imaging, it is usually necessary to collect a large amount of underwater data, which will bring great challenges to the corresponding hardware devices and communication systems that send measurement data. Therefore, it is of great practical significance to study how to reduce the amount of data required while ensuring the imaging quality. Compression sensing theory (Compressed Sensing) can reduce the minimum sampling frequency of distorted reconstructed signal to less than Nyquist frequency, which can be used to solve the problem of large amount of imaging data. In this paper, two synthetic aperture imaging algorithms using compression sensing are studied. Since the sparse matrix of sonar signal is located in the complex field, this paper presents the realization form of the complex domain of reconstruction algorithm. Finally, the performance of the algorithm is verified by the imaging simulation of point target. The main work of this paper is as follows: (1) the basic theory and basic knowledge of compression sensing are introduced in detail, including sparse signal, nonlinear measurement and reconstruction of original signal. The principle of synthetic aperture sonar imaging and the classical range Doppler imaging algorithm are studied, and the imaging simulation of the four-point target is completed. (2) in order to solve the problem of large amount of data in range direction, the imaging algorithm of range compression sensing with phase-preserving property is studied. The experimental results show that the algorithm can obtain a good imaging effect with only 20% of the original data in the distance direction, and the algorithm itself can suppress the sidelobe and has a good anti-noise performance. In order to obtain longer observation distance without azimuth ambiguity, a two-dimensional downsampling compression sensing imaging algorithm is studied, which can further reduce the amount of data while guaranteeing the imaging quality. (3) the complex domain reconstruction of two compression sensing imaging algorithms is discussed, that is, the decomposed complex domain reconstruction form of classical algorithm and the SPGL1 algorithm which can be solved directly in complex field. The results show that, SPGL1 algorithm not only can save storage space, but also has good anti-noise performance.
【学位授予单位】：大连理工大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：P715.5
本文编号：2339600
[Abstract]:With the rapid development of marine industry in China, the requirement of underwater imaging quality is increasing, and synthetic aperture sonar equipment with high resolution imaging capability has gradually become a research hotspot. In order to obtain high resolution imaging, it is usually necessary to collect a large amount of underwater data, which will bring great challenges to the corresponding hardware devices and communication systems that send measurement data. Therefore, it is of great practical significance to study how to reduce the amount of data required while ensuring the imaging quality. Compression sensing theory (Compressed Sensing) can reduce the minimum sampling frequency of distorted reconstructed signal to less than Nyquist frequency, which can be used to solve the problem of large amount of imaging data. In this paper, two synthetic aperture imaging algorithms using compression sensing are studied. Since the sparse matrix of sonar signal is located in the complex field, this paper presents the realization form of the complex domain of reconstruction algorithm. Finally, the performance of the algorithm is verified by the imaging simulation of point target. The main work of this paper is as follows: (1) the basic theory and basic knowledge of compression sensing are introduced in detail, including sparse signal, nonlinear measurement and reconstruction of original signal. The principle of synthetic aperture sonar imaging and the classical range Doppler imaging algorithm are studied, and the imaging simulation of the four-point target is completed. (2) in order to solve the problem of large amount of data in range direction, the imaging algorithm of range compression sensing with phase-preserving property is studied. The experimental results show that the algorithm can obtain a good imaging effect with only 20% of the original data in the distance direction, and the algorithm itself can suppress the sidelobe and has a good anti-noise performance. In order to obtain longer observation distance without azimuth ambiguity, a two-dimensional downsampling compression sensing imaging algorithm is studied, which can further reduce the amount of data while guaranteeing the imaging quality. (3) the complex domain reconstruction of two compression sensing imaging algorithms is discussed, that is, the decomposed complex domain reconstruction form of classical algorithm and the SPGL1 algorithm which can be solved directly in complex field. The results show that, SPGL1 algorithm not only can save storage space, but also has good anti-noise performance.
【学位授予单位】：大连理工大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：P715.5
【参考文献】
相关期刊论文 前4条
1 李军,霍国正,郑一鸣;合成孔径声纳新进展[J];舰船电子工程;2000年06期
2 ZHAO Yao;FENG Jing;ZHANG BingChen;HONG Wen;WU YiRong;;Current progress in sparse signal processing applied to radar imaging[J];Science China(Technological Sciences);2013年12期
3 ;A novel spaceborne SAR wide-swath imaging approach based on Poisson disk-like nonuniform sampling and compressive sensing[J];Science China(Information Sciences);2012年08期
4 张春华;刘纪元;;第二讲 合成孔径声纳成像及其研究进展[J];物理;2006年05期
，本文编号：2339600
本文链接：https://www.wllwen.com/kejilunwen/haiyang/2339600.html
