中文摘要：

声传感器基阵式成像是水下目标探测的主要方法之一。在保证成像质量的同时,采用稀疏基阵是降低系统复杂性的有效途径。另一方面,成像方法也是关键技术之一,其性能直接决定成像质量的优劣。在前期研究的稀疏阵基础上,提出了压缩感知成像方法用于水下成像。回波经传感器基阵接收并进行信号分离后,获得各虚拟通道信号,利用压缩感知方法进行成像。在提出的成像方法中,采用了自适应块贝叶斯算法,它能够在噪声环境下较为准确地恢复图像。此外,根据声信号特点设计字典,并按算法特性设计了数据排列规则。实验表明：通过自适应块贝叶斯算法的压缩感知成像能够可靠地对水下目标进行成像,目标的几何特征明显。

英文摘要：

The acoustic sensor array based imaging is one of the major ways to be used to find the underwater targets. While ensuring the imaging quality, adopting sparse array is the effective way to reduce the imaging system complexity. On the other hand, the imaging method is also one of the key technologies in the imaging, its performance directly decides the quality and rate of imaging. On the basis of previous study on sparse array, in this paper the compressive sensing imaging method is proposed and used in the underwater acoustic imaging. The echo signals are received by the sparse array, the signal separation is implemented and the virtual channel signals are obtained; the compressive sensing algorithm is used to obtain the image. In the proposed imaging method, the adaptive block Bayes algorithm is adopted, which can precisely recover the sparse target image in noise environment. Moreover, a dictionary is reasonably designed according to the features of the acoustic signals, and the data matrix of the separated signals is rearranged according to the characteristics of the algorithm to facilitate the algorithm. The experiment results show that with the compressive sensing imaging method based on adaptive block Bayes algorithm, the underwater target imaging can be implemented reliably, and the complete geometrical characteristics of the underwater target can be clearly displayed.