Search

In synthetic aperture imaging, the interpulse Doppler shift is commonly used, while the intrapulse Doppler shift is ignored. However, for the linear frequency modulation signal, the intrapulse Doppler shift can cause an additional range cell migration (ARCM) and a phase error term. Additionally, the ARCM may exceed a range resolution cell in certain cases of multireceiver synthetic aperture sonar (SAS), which will result in image defocus. Therefore, the intrapulse Doppler shift in multireceiver SAS should not be ignored in such cases. In this study, the condition of the negligible intrapulse Doppler shift is presented, that is, the moving distance of the sonar platform within a pulse should be much less than an azimuth resolution cell. Based on the multireceiver accurate time-delay model, a multireceiver SAS intrapulse Doppler shift signal model is developed by introducing the intrapulse Doppler frequency into the echo signal. After simplifying the accurate range history under the narrow-beam assumption, the range-Doppler spectrum is derived. Finally, a multireceiver SAS chirp scaling algorithm considering the intrapulse Doppler shift (IDS-CSA) is derived. The proposed algorithm can correct the ARCM and compensate the phase error (although not required), and then obtain a well-focused synthetic aperture image. The effectiveness of the proposed model and algorithm is verified by simulations and field data. The results show that the proposed algorithm provides a better image quality than the conventional multireceiver CSA when the condition of the negligible intrapulse Doppler shift is not met.

Recent deep learning (DL) detectors adopted by radar or sonar (RS) are normally trained with transfer learning, where the typical workflow is to pretrain a convolutional neural network (CNN) on external large-scale classification datasets (e.g., ImageNet) as the backbone and then finetune the entire detector on detection datasets. Though transfer learning could effectively avoid overfitting, transferred models are usually redundant and might not generalize well on RS datasets. To achieve high generalization and to eliminate the dependence on transfer learning, a self-trained target detection method is established by including Automatic Deep Learning (AutoDL) to design optimal detectors. This self-trained target detection consists of three stages. First, a derived classification dataset (DCD) consisting of image blocks of targets and backgrounds is derived from detection datasets. Then, a memory-efficient Differentiable Architecture Search algorithm with flexible search space and large inputs (FL-DARTS), which is characterized by its predefined multistride convolutions, poolings, and unique super-structure, is proposed to automatically design and self-train optimal CNNs on DCDs. Finally, self-trained AutoDL detectors are implemented with the automatic backbone designed by FL-DARTS. We evaluated three self-trained AutoDL detectors on the public SAR ship detection dataset (SSDD) and the self-made sonar common target detection dataset (SCTD). The experiments show that while the number of parameters of automatic backbones designed for SSDD and SCTD are only 11.8% and 15.2% of that of ResNet50, self-trained AutoDL detectors implemented with automatic backbones significantly outperform their transfer learning detectors and achieve state-of-the-art detection precisions and high detection speeds. Data, codes are publicly available.

Two-dimensional phase unwrapping is a crucial step in interferometric signal processing. A phase quality map can help the unwrapping algorithm deal with low-quality and fast-changing regions. However, because existing algorithms cannot calculate a quality map representing the gradient quality directly, it is usually necessary to approximate the gradient quality with phase quality to assist the network-based phase unwrapping algorithm. Furthermore, they cannot withstand intense noise in low-quality regions, resulting in many errors in path-based algorithms. To address the aforementioned issues, this paper analyzes the essence of a quality map and proposes a quality map generation method based on a convolutional neural network. The generated quality maps are a pair, each indicating the quality of horizontal and vertical gradients. Experiments show that the quality map generated by this method can help path-based and network-based algorithms perform better.

The existing imaging algorithms for the multiple-hydrophone synthetic aperture sonar (SAS) cannot obtain a good imaging result in the moderate squint case where the squint angle is from 10° to 20°, because they do not take the differential range curvature into account, and the existing methods of equivalent spatial sampling cannot work well. This study proposes an imaging algorithm for the moderate squint multiple-hydrophone SAS. First, the range history model of the moderate squint multiple-hydrophone SAS is established. Second, the signal of each hydrophone is processed separately with bistatic imaging algorithm to correct the differential range curvature. Third, the corrected signal of each hydrophone is shifted to the same range bin to compensate the relative range offset, and is added coherently together to remove the azimuth aliasing and to reconstruct the image without loss of resolution. Fourth, a new method of equivalent azimuth sampling is developed, and a PRF adjusted with squint angle is given. Finally, the validity of the proposed algorithm is tested by the simulation and real data.

A hierarchical quality-guided phase unwrapping algorithm in a shared memory environment is proposed. First, the wrapped phase is divided into regular blocks, and local wrap counts of every block are obtained by a quality-guided strategy. Then, the gradient of the block wrap counts of adjacent blocks and the quality of every block are defined by the boundary local wrap counts of every block. Each block's data can be regarded as an abstract phase point, and the quality-guided strategy can be used again to solve the block wrap counts of each block. Finally, the absolute wrap counts of each phase point are obtained by adding local wrap counts and corresponding block wrap counts, and then the final unwrapped phase is obtained. The performance of the proposed algorithm is verified through an unwrapping experiment performed on simulated data and the real interferometric synthetic aperture sonar wrapped phase in a shared memory environment. The results show that the proposed method greatly improves phase unwrapping efficiency while maintaining the correctness of unwrapped results.

For the imaging of multiple-receiver synthetic aperture sonar (SAS) under the nonstop-hop-stop case, an exact analytical solution for the two-dimensional (2-D) frequency spectrum is usually difficult to obtain because of the existence of the double-square-root (DSR) term in the range history equation. Several approximate solutions for the 2-D spectrum have been derived to focus the multiple-receiver SAS data. In this paper, the range history geometry for multiple-receiver SAS is newly constructed to meet the demand of the derivation of an analytical 2-D spectrum. According to the geometry-based bistatic formula (GBF) method, a quasi-analytical 2-D spectrum with an unknown variable named half quasi-bistatic angle (HQBA) is reviewed. Based on the method of series reversion (MSR), the fourth order equation with respect to the HQBA is solved and an analytical HQBA is obtained. After substituting the analytical HQBA into the quasi-analytical 2-D spectrum, a completely analytical 2-D spectrum is obtained and the problem of the 2-D spectrum derivation caused by the DSR term is solved successfully. In this paper, an extended range Doppler (RD) algorithm based on the derived 2-D spectrum is proposed for focusing the multiple-receiver SAS data under the non stop-hop-stop case. The results of simulation and real experiment data imaging have validated the effectiveness and accuracy of the proposed algorithm.

A correction approach for the inclined array of hydrophones is proposed to prevent decline of the image quality in SAS. In this approach, the 2-way exact acoustic propagation path of the inclined array is transformed into the sum of a single root term and an offset term, where the single root term is the 2-way ideal propagation path and the offset term contains all errors cause by the inclined array. The correction for the offset term is separated into two steps: phase correction and delay correction. The phase correction is performed on the echo signal of each receiving hydrophone in the 2-D time domain by a phase multiplication and the delay correction is performed on the echo signal of each receiving hydrophone in the range frequency domain by a phase multiplication with a linear function of range frequency at the reference range. Finally, the effectiveness of the proposed approach is examined by the simulation experiments.

Phase unwrapping is one of the key problems in reconstructing the digital elevation model of a scene from its interferometric synthetic aperture radar (InSAR) or interferometric synthetic aperture sonar (InSAS) data. In this letter, we propose a quality-guided and local minimum discontinuity based phase unwrapping algorithm, which enhances the precision of the unwrapped result by local minimum discontinuity optimization in low quality areas, and still keeps a high efficiency. The new algorithm can be divided into two steps. Firstly, the quality-guided phase unwrapping algorithm is performed on the original wrapped phase image, which helps to remove all the discontinuities caused by the wrapping operation quickly, but tends to spread the unwrapped errors in low quality areas. Secondly, the initial unwrapped result is divided into high and low quality areas according to its quality map, and the minimum discontinuity optimization process is performed in the low quality areas of the initial unwrapped result, which helps to remove the remaining improving loops defined as a loop with more positive jumps than negative ones. This prevents the unwrapped errors spreading from low quality areas to high quality areas and accelerates the optimization process by restricting the optimization place in low quality areas. Tests performed on InSAR and real InSAS data confirm the accuracy and efficiency of the proposed algorithm.

In the above letter (ibid., vol. 8, no. 2, pp. 364-368, Mar. 2011), in the Abstract, a sentence is incorrect. The correct sentence is presented here.