Active sonar plays an essential role in the development of the marine industry. The elements of the transmitting transducer array may be damaged due to prolonged use or harsh working conditions, causing a severe decline in performance. Due to the high complexity and low signal-to-noise ratio (SNR), the reliable and nondestructive diagnosis of failed elements in the underwater environment is challenging. This article proposes the acoustic image reconstruction diagnostic methodology (AIRDM) aimed at diagnosing array element failures. The method collects multidirectional transmitted signals through one hydrophone to reconstruct the acoustic image of the array, and failed elements can be identified. In addition, a time-delay correction is incorporated to achieve diagnosis in the near-field condition, which extends the range of applications. The simulation and actual experiment results demonstrate that the AIRDM is more accurate and reliable at low SNR than the existing antenna field diagnostic methods. This study will effectively promote the reliable operation of linear transmitting arrays, such as sonar and antenna.

Beamforming is extensively used in underwater acoustic imaging systems. As a high-resolution variety, a Bayesian compressive beamformer treats the acoustic echo of each snapshot independently, and achieves enhanced recovery performance. However, its associated computational cost is extremely high compared with conventional beamformers, and its iterative implementation makes it difficult to be used online. To overcome these obstacles, an online Bayesian compressive beamformer based on Kalman filtering (online-KSBL) is proposed in this work, which is non-iterative and computationally efficient for a continual working scenario, where a long-term imaging task is conducted underwater. Imposing independent assumptions on snapshots, the online-SBL approach can be derived from online-KSBL. The hyperparameters of the model are estimated recursively, following efficient procedures that are implemented approximately with a sawtooth lag scheme. For the underwater imaging scenario, the correlation among snapshots is exploited and inferred from online-KSBL, and it is found that online-SBL performs competitively with online-KSBL, and it is sufficient to employ online-SBL without considering the snapshot correlation, retaining low complexity, as demonstrated by experimental results.

This letter presents a multi-target tracking algorithm—thresholded Sequential Monte Carlo probability hypothesis density (TH-SMC-PHD) algorithm. The TH-SMC-PHD aims to overcome the problem that underwater multi-target tracking is prone to missing tracking on sonar images, resulting in the breakage of trajectories. First, CA-CFAR and K-means are employed to detect potential underwater targets from sonar images, respectively. Then TH-SMC-PHD is applied to the detection results for tracking, using a continuously lost frame threshold to reduce the missing tracking rate and the Minimum-Sampling-Variance (MSV) resampling to improve tracking accuracy. An actual underwater multi-target tracking experiment using a forward-looking sonar was contracted in a water tank to evaluate the tracking performance. Compared with the other two PHD tracking algorithms, the results demonstrate that the proposed algorithm achieves high-precision, non-fracture, non-missing tracking of three targets. In addition, the TH-SMC-PHD is more stable and less affected by different detection algorithms, which has the potential for practical applications.

Forward-looking sonar is one of the essential imaging equipment employed in exploring underwater targets. However, it is always challenging to detect targets from sonar images considering the complex environment. This paper represents an automatic underwater target detection method using clustering, segmentation, and feature discrimination. Firstly, we combine the Fuzzy C-means Clustering (FCM) and K-means to cluster the sonar image globally to obtain as many Regions of Interests (ROIs) as possible. Secondly, the Pulse Coupled Neural Network (PCNN) is used to locally segment the target boundary from the ROIs. Finally, multiple features are extracted from the target area as the feature vector, which is inputted into the nonlinear converter to enlarge the features’ distance. Then we use Fisher discriminant to estimate the classification threshold, which realizes the underwater target detection. The experimental results show that the proposed method has low detection error and good real-time performance under low false alarm probability, which is not inferior to the popular deep learning approaches at present.

Compressive sensing (CS) is a promising technique recognized for its merits in recovering sparse signals with enhanced resolution entailing specific constraints. In recasting the CS model within a Bayesian framework, its formulation can be interpreted and solved under various prior assumptions that may correspond to, e.g., the$\mathcal {L}_{1}$or reweighted$\mathcal {L}_{1}$constraint. Bayesian CS (or Bayesian sparse learning, BSL) achieves improved resolution and robustness compared with the deterministic CS. Therefore, BSL has been invoked to solve the single and multisnapshot beamforming models for direction-of-arrival (DOA) estimation. However, the recovery performance deteriorates for complicated signals because of nonsparsity. In this article, a multiconstraint BSL approach is proposed to solve the multisnapshot beamforming model (termed M-MCRBSL), which reconstructs the amplitude and DOA of the source simultaneously. With a proper assembly of constraints, the source can be represented sparsely and the transformation coefficients can be recovered accurately in multiple sparse domains. As demonstrated in the simulations, M-MCRBSL outperforms other state-of-the-art multisnapshot beamforming methods gauged by both the normalized mean square error (NMSE) and the structural similarity (SSIM) index. In particular, a deficiency sensitivity experiment is devised to elaborate the feasibility of the invoked invertibility approximation. As attested with an underwater acoustics experiment, M-MCRBSL with joint identity and Haar wavelet constraints achieves improved performance in terms of enhanced resolution and suppressed interference sources.

We study the cluster ensemble problem and propose a cluster ensemble approach based on subspace similarity (CEASS). From a subspace similarity perspective, we seek the optimal subspace which is most similar to the given subspaces corresponding to the cluster solutions to be combined. We formulate t he cluster ensemble problem as an optimization problem of minimizing the squared sum of Euclidean distances between the standard orthogonal basis vectors of the target subspace and the given subspaces. We derived an explicit solution to the preceding problem in terms of singular value decomposition. Moreover, the solution consists of the low dimensional embeddings of instances. Finally, K-means algorithm with the minimum-maximum principle is utilized to cluster instances according to their coordinates in the embedding space. In particular, we circumvent the initialization problem of K-means by employing CEASS that combines different K-means clustering solutions obtained from random initialization to obtain a stable clustering result. We evaluate and compare CEASS so constructed with several other state-of-art cluster ensemble algorithms using nine real world datasets. Experimental results demonstrate that CEASS generally outperforms other algorithms in terms of normalized mutual information and F1 measure. In addition, CEASS is extremely efficient compared to hierarchy clustering algorithms.