Explore the vast range of titles available.

Sonar is a valuable tool for ocean exploration since it can obtain a wealth of data. With the development of intelligent technology, deep learning has brought new vitality to underwater sonar image classification. However, due to the difficulty and high cost of acquiring underwater sonar images, we have to consider the extreme case when there are no available sonar data of a specific category, and how to improve the prediction ability of intelligent classification models for unseen sonar data. In this work, we design an underwater sonar image classification method based on Image Disentanglement Reconstruction and Zero-Shot Learning (IDR-ZSL). Initially, an image disentanglement reconstruction (IDR) network is proposed for generating pseudo-sonar samples. The IDR consists of two encoders, a decoder, and three discriminators. The first encoder is responsible for extracting the structure vectors of the optical images and the texture vectors of the sonar images; the decoder is in charge of combining the above vectors to generate the pseudo-sonar images; and the second encoder is in charge of disentangling the pseudo-sonar images. Furthermore, three discriminators are incorporated to determine the realness and texture quality of the reconstructed image and feedback to the decoder. Subsequently, the underwater sonar image classification model performs zero-shot learning based on the generated pseudo-sonar images. Experimental results show that IDR-ZSL can generate high-quality pseudo-sonar images, and improve the prediction accuracy of the zero-shot classifier on unseen classes of sonar images.

Deep neural network (DNN) was applied in sonar image target recognition tasks, but it is very difficult to obtain enough sonar images that contain a target; as a result, the direct use of a small amount of data to train a DNN will cause overfitting and other problems. Transfer learning is the most effective way to address such scenarios. However, there is a large domain gap between optical images and sonar images, and common transfer learning methods may not be able to effectively handle it. In this paper, we propose a transfer learning method for sonar image classification and object detection called the texture feature removal network. We regard the texture features of an image as domain-specific features, and we narrow the domain gap by discarding the domain-specific features, and hence, make it easier to complete knowledge transfer. Our method can be easily embedded into other transfer learning methods, which makes it easier to apply to different application scenarios. Experimental results show that our method is effective in side-scan sonar image classification tasks and forward-looking sonar image detection tasks. For side-scan sonar image classification tasks, the classification accuracy of our method is enhanced by 4.5% in a supervised learning experiment, and for forward-looking sonar detection tasks, the average precision (AP) is also significantly improved.

To overcome the shortcomings of the traditional manual detection of underwater targets in side-scan sonar (SSS) images, a real-time automatic target recognition (ATR) method is proposed in this paper. This method consists of image preprocessing, sampling, ATR by integration of the transformer module and YOLOv5s (that is, TR–YOLOv5s), and target localization. By considering the target-sparse and feature-barren characteristics of SSS images, a novel TR–YOLOv5s network and a down-sampling principle are put forward, and the attention mechanism is introduced in the method to meet the requirements of accuracy and efficiency for underwater target recognition. Experiments verified the proposed method achieved 85.6% mean average precision (mAP) and 87.8% macro-F2 score, and brought 12.5% and 10.6% gains compared with the YOLOv5s network trained from scratch, and had the real-time recognition speed of about 0.068 s per image.

Forward-looking sonar is a technique widely used for underwater detection. However, most sonar images have underwater noise and low resolution due to their acoustic properties. In recent years, the semantic segmentation model U-Net has shown excellent segmentation performance, and it has great potential in forward-looking sonar image segmentation. However, forward-looking sonar images are affected by noise, which prevents the existing U-Net model from segmenting small objects effectively. Therefore, this study presents a forward-looking sonar semantic segmentation model called Feature Pyramid U-Net with Attention (FPUA). This model uses residual blocks to improve the training depth of the network. To improve the segmentation accuracy of the network for small objects, a feature pyramid module combined with an attention structure is introduced. This improves the model’s ability to learn deep semantic and shallow detail information. First, the proposed model is compared against other deep learning models and on two datasets, of which one was collected in a tank environment and the other was collected in a real marine environment. To further test the validity of the model, a real forward-looking sonar system was devised and employed in the lake trials. The results show that the proposed model performs better than the other models for small-object and few-sample classes and that it is competitive in semantic segmentation of forward-looking sonar images.

Due to the strong speckle noise caused by the seabed reverberation which makes it difficult to extract discriminating and noiseless features of a target, recognition and classification of underwater targets using side-scan sonar (SSS) images is a big challenge. Moreover, unlike classification of optical images which can use a large dataset to train the classifier, classification of SSS images usually has to exploit a very small dataset for training, which may cause classifier overfitting. Compared with traditional feature extraction methods using descriptors—such as Haar, SIFT, and LBP—deep learning-based methods are more powerful in capturing discriminating features. After training on a large optical dataset, e.g., ImageNet, direct fine-tuning method brings improvement to the sonar image classification using a small-size SSS image dataset. However, due to the different statistical characteristics between optical images and sonar images, transfer learning methods—e.g., fine-tuning—lack cross-domain adaptability, and therefore cannot achieve very satisfactory results. In this paper, a multi-domain collaborative transfer learning (MDCTL) method with multi-scale repeated attention mechanism (MSRAM) is proposed for improving the accuracy of underwater sonar image classification. In the MDCTL method, low-level characteristic similarity between SSS images and synthetic aperture radar (SAR) images, and high-level representation similarity between SSS images and optical images are used together to enhance the feature extraction ability of the deep learning model. Using different characteristics of multi-domain data to efficiently capture useful features for the sonar image classification, MDCTL offers a new way for transfer learning. MSRAM is used to effectively combine multi-scale features to make the proposed model pay more attention to the shape details of the target excluding the noise. Experimental results of classification show that, in using multi-domain data sets, the proposed method is more stable with an overall accuracy of 99.21%, bringing an improvement of 4.54% compared with the fine-tuned VGG19. Results given by diverse visualization methods also demonstrate that the method is more powerful in feature representation by using the MDCTL and MSRAM.

Target detection in side-scan sonar images plays a significant role in ocean engineering. However, the target images are usually severely interfered by the complex background and strong environmental noise, which makes it difficult to extract robust features from small targets and makes the target detection task quite challenging. In this paper, a novel small target detection method in sonar images is proposed based on the low-rank sparse matrix factorization. Initially, the side-scan sonar images are preprocessed so as to highlight the individual differences of the target. Then, the problems of target feature extraction and noise removal are characterized as the problem of matrix decomposition. An improved Robust Principal Component Analysis algorithm is used to extract target information, and the fast proximal gradient method is used to optimize the solution. The original sonar image is reconstructed into the low-rank background matrix, the sparse target matrix, and the noise matrix. Eventually, a morphological operation is used to filter out the noise and refine the target edges in the target matrix for improving the accuracy of target detection. Experimental results show that the proposed method not only achieves better detection performance in comparison to the conventional baseline algorithms but also performs robustly in various signal-to-clutter ratio conditions.

Sonar image classification is crucial in salvage operations and submarine pipeline detection. However, it faces challenges of low resolution, few-shot, and long-tail due to multipath interference and data collection issues. Current methods employ transfer learning, resampling, and adversarial attacks to address these challenges. Nonetheless, knowledge transfer from optical to sonar images is often inefficient due to significant domain differences. Furthermore, the large receptive fields of existing models make it difficult to extract local details from low-resolution sonar images. Additionally, cross-entropy loss excessively suppresses tail class gradients, causing a bias towards head classes. To address these problems, this paper proposes a Hierarchical Transfer Progressive Learning based on the Jigsaw puzzle and Block Convolution (HTPL-JB). First, we introduce a hierarchical pre-training strategy incorporating a source pre-training phase into the transfer learning phase, enhancing the efficiency of transferring knowledge from optical to sonar images. In the fine-tuning phase, we employ a progressive training strategy to progressively extract information at different granular levels, enhancing the model’s ability to capture fine details from sonar images. Finally, we introduce a key point sensitive loss (KPSLoss), which uses a larger margin distance and a smaller slope factor for the tail class to enhance accuracy and the separability of key points. Extensive experiments on the NKSID datasets demonstrate that HTPL-JB significantly outperforms the existing methods. Our code will be available at https://github.com/leeAndJim/JBHTPL .

Due to the small sample size of underwater acoustic data and the strong noise interference caused by seabed reverberation, recognizing underwater targets in Side-Scan Sonar (SSS) images is challenging. Using a transfer-learning-based recognition method to train the backbone network on a large optical dataset (ImageNet) and fine-tuning the head network with a small SSS image dataset can improve the classification of sonar images. However, optical and sonar images have different statistical characteristics, directly affecting transfer-learning-based target recognition. In order to improve the accuracy of underwater sonar image classification, a style transformation method between optical and SSS images is proposed in this study. In the proposed method, objects with the SSS style were synthesized through content image feature extraction and image style transfer to reduce the variability of different data sources. A staged optimization strategy using multi-modal data effectively captures the anti-noise features of sonar images, providing a new learning method for transfer learning. The results of the classification experiment showed that the approach is more stable when using synthetic data and other multi-modal datasets, with an overall accuracy of 100%.

Side scan sonar images have great application prospects in underwater surveys, target detection, and engineering activities. However, the acquired sonar images exhibit low illumination, scattered noise, distorted outlines, and unclear edge textures due to the complicated undersea environment and intrinsic device flaws. Hence, this paper proposes a multi-scale fusion strategy for side scan sonar (SSS) image correction to improve the low contrast and noise interference. Initially, an SSS image was decomposed into low and high frequency sub-bands via the non-subsampled shearlet transform (NSST). Then, modified multi-scale retinex (MMSR) was employed to enhance the contrast of the low frequency sub-band. Next, sparse dictionary learning (SDL) was utilized to eliminate high frequency noise. Finally, the process of NSST reconstruction was completed by fusing the emerging low and high frequency sub-band images to generate a new sonar image. The experimental results demonstrate that the target features, underwater terrain, and edge contours could be clearly displayed in the image corrected by the multi-scale fusion strategy when compared to eight correction techniques: BPDHE, MSRCR, NPE, ALTM, LIME, FE, WT, and TVRLRA. Effective control was achieved over the speckle noise of the sonar image. Furthermore, the AG, STD, and E values illustrated the delicacy and contrast of the corrected images processed by the proposed strategy. The PSNR value revealed that the proposed strategy outperformed the advanced TVRLRA technology in terms of filtering performance by at least 8.8%. It can provide sonar imagery that is appropriate for various circumstances.

Sonar images are inherently affected by speckle noise, which degrades image quality and hinders image exploitation. Despeckling is an important pre-processing task that aims to remove such noise so as to improve the accuracy of analysis tasks on sonar images. In this paper, we propose a novel transformer-based generative adversarial network named SID-TGAN for sonar image despeckling. In the SID-TGAN framework, transformer and convolutional blocks are used to extract global and local features, which are further integrated into the generator and discriminator networks for feature fusion and enhancement. By leveraging adversarial training, SID-TGAN learns more comprehensive representations of sonar images and shows outstanding performance in speckle denoising. Meanwhile, SID-TGAN introduces a new adversarial loss function that combines image content, local texture style, and global similarity to reduce image distortion and information loss during training. Finally, we compare SID-TGAN with state-of-the-art despeckling methods on one image dataset with synthetic optical noise and four real sonar image datasets. The results show that it achieves significantly better despeckling performance than existing methods on all five datasets.