Sidescan sonars are increasingly used in underwater search and rescue for drowning victims, wrecks and airplanes. Automatic object classification or detection methods can help a lot in case of long searches, where sonar operators may feel exhausted and therefore miss the possible object. However, most of the existing underwater object detection methods for sidescan sonar images are aimed at detecting minelike objects, ignoring the classification of civilian objects, mainly due to lack of dataset. So, in this study, we focus on the multi-class classification of drowning victim, wreck, airplane, mine and seafloor in sonar images. Firstly, through a long-term accumulation, we built a real sidescan sonar image dataset named SeabedObjects-KLSG, which currently contains 385 wreck, 36 drowning victim, 62 airplane, 129 mine and 578 seafloor images. Secondly, considering the real dataset is imbalanced, we proposed a semisynthetic data generation method for producing sonar images of airplanes and drowning victims, which uses optical images as input, and combines image segmentation with intensity distribution simulation of different regions. Finally, we demonstrate that by transferring a pre-trained deep convolutional neural network (CNN), e.g. VGG19, and fine-tuning the deep CNN using 70% of the real dataset and the semisynthetic data for training, the overall accuracy on the remaining 30% of the real dataset can be eventually improved to 97.76%, which is the highest among all the methods. Our work indicates that the combination of semisynthetic data generation and deep transfer learning is an effective way to improve the accuracy of underwater object classification.

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.

To better solve the problem of target detection in marine environment and to deal with the difficulty of 3D reconstruction of underwater target, a binocular vision-based underwater target detection and 3D reconstruction system is proposed in this paper. Two optical sensors are used as the vision of the system. Firstly, denoising and color restoration are performed on the image sequence acquired by the vision of the system and the underwater target is segmented and extracted according to the image saliency using the super-pixel segmentation method. Secondly, aiming to reduce mismatch, we improve the semi-global stereo matching method by strictly constraining the matching in the valid target area and then optimizing the basic disparity map within each super-pixel area using the least squares fitting interpolation method. Finally, based on the optimized disparity map, triangulation principle is used to calculate the three-dimensional data of the target and the 3D structure and color information of the target can be given by MeshLab. The experimental results show that for a specific size underwater target, the system can achieve higher measurement accuracy and better 3D reconstruction effect within a suitable distance.

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%.

Existing size measurement systems cannot meet the requirements of non-contact measurement tasks, and the main challenge is how to detect various objects and improve measure accuracy. In order to solve these problems, a novel measurement system was proposed in this paper. In the system, a three-camera model with variable baselines was designed based on the trinocular vision. Three binocular vision subsystems were composed of the three cameras, which were used to obtain depth information from different shooting angles, and the baselines between cameras could be adjusted according to different objects. In measurement process, target object was detected automatically based on visual saliency features and spatial information. Finally, the size of target object was computed by cooperative analysis of the three binocular vision subsystems. The experimental results demonstrated that the proposed system is accurate and stable in various objects detection and measurement tasks.

For synthetic aperture radar (SAR) images, traditional edge detection methods can hardly extract the complete and true edges since they are sensitive to noise. In this study, the authors propose an edge detection method based on bionic vision in non-subsampled contourlet transform (NSCT) domain for SAR images, which makes use of the characteristics of NSCT (e.g. multiscale, multidirection edge expression and outstanding edge location) and the mechanisms of strengthening and positioning the image edge by fixational eye movements. First, source images are decomposed in the NSCT domain. The minimum mean squared error (MMSE) filter in the NSCT domain is adopted to reduce speckle noise of SAR images. Movement, difference and competition are applied on each subband to obtain original multiscale and multidirection subband edges. Lipschitz regularity and non-maximum suppression are also adopted to effectively remove noise points. Finally, hysteresis thresholding with adaptive thresholds is introduced to avoid streaking and trial-and-error, and multiscale edge fusion is used to reduce the false alarm rate. Edge extraction results of simulated and real SAR images demonstrate that the proposed method is better than other edge detection methods based on Canny operator, ratio of averages (ROA) operator, ratio of exponentially weighted averages (ROEWA) operator, wavelet modulus maximum and curvelet.