The demand for fish products, which provide crucial protein for humans, is rising as the global population grows. In contrast, fish stock is declining due to human activity, environmental changes and ...overfishing. Fish monitoring provides valuable support data for effective fishery management and ecosystem conservation. The common monitoring methods are based on manual sampling, which is time‐consuming, laborious and intrusive. Imaging sonar is a hydroacoustic system that produces acoustic images similar to optical images by transmitting and receiving sound waves, allowing for in situ monitoring of fish non‐intrusively in the dark and turbid water environments where optical cameras are limited. In the last decade, imaging sonar, especially high frequency multibeam forward‐looking sonar and side‐scan sonar, has been widely used in fish monitoring. We reviewed the literature from the previous decade on the use of these two types of imaging sonar in fish species identification, abundance estimation, length measurement and behaviour analysis, as well as the sonar imagery processing concerning fish. The review results show that these imaging sonars are efficient and effective tools for fish monitoring in complex environments. The challenges include (1) the recognition of small fish forming dense aggregations; (2) species identification, which limits their use in species‐specific studies; (3) time‐consuming massive data processing. Therefore, advanced algorithms for sonar imagery processing and integrations with other sampling technologies are needed for future development.
Underwater acoustics has been implemented mostly in the field of sound navigation and ranging (SONAR) procedures for submarine communication, the examination of maritime assets and environment ...surveying, target and object recognition, and measurement and study of acoustic sources in the underwater atmosphere. With the rapid development in science and technology, the advancement in sonar systems has increased, resulting in a decrement in underwater casualties. The sonar signal processing and automatic target recognition using sonar signals or imagery is itself a challenging process. Meanwhile, highly advanced data-driven machine-learning and deep learning-based methods are being implemented for acquiring several types of information from underwater sound data. This paper reviews the recent sonar automatic target recognition, tracking, or detection works using deep learning algorithms. A thorough study of the available works is done, and the operating procedure, results, and other necessary details regarding the data acquisition process, the dataset used, and the information regarding hyper-parameters is presented in this article. This paper will be of great assistance for upcoming scholars to start their work on sonar automatic target recognition.
Traditional chirp scaling algorithm is often developed based on the quadratic approximation of point target reference spectrum, which neglects the higher‐order coupling between range and azimuth ...dimensions. Besides, the scaling operation is conducted based on quadratic term of reference range, and the residual quadratic error increases with target range far away from reference range. The analysis of phase error shows that both errors are noticeable under the case of wide‐bandwidth signal, and the imaging performance would be seriously distorted. To solve this problem, the higher‐order coupling and quadratic error are compensated by exploiting sub‐block processing method in the range dimension when the chirp scaling algorithm is exploited to reconstruct targets. Compared to traditional methods, the presented method significantly improves the focussing performance based on the simulation results. The processing of real data further shows that the performance of presented method is superior to that of traditional method.
Due to the complexity of the ocean environment, an autonomous underwater vehicle (AUV) is disturbed by obstacles when performing tasks. Therefore, the research on underwater obstacle detection and ...avoidance is particularly important. Based on the images collected by a forward-looking sonar on an AUV, this article proposes an obstacle detection and avoidance algorithm. First, a deep learning-based obstacle candidate area detection algorithm is developed. This algorithm uses the You Only Look Once (YOLO) v3 network to determine obstacle candidate areas in a sonar image. Then, in the determined obstacle candidate areas, the obstacle detection algorithm based on the improved threshold segmentation algorithm is used to detect obstacles accurately. Finally, using the obstacle detection results obtained from the sonar images, an obstacle avoidance algorithm based on deep reinforcement learning (DRL) is developed to plan a reasonable obstacle avoidance path of an AUV. Experimental results show that the proposed algorithms improve obstacle detection accuracy and processing speed of sonar images. At the same time, the proposed algorithms ensure AUV navigation safety in a complex obstacle environment.
The covertness of an active sonar is a very important issue and the sonar signal waveform design problem is studied to improve covertness of the system. Many marine mammals produce call pulses for ...communication and echolocation, and existing interception systems normally classify these biological signals as ocean noise and filter them out. Based on this, a disguised sonar signal waveform design approach with its camouflage application strategy for underwater sensor platforms is proposed by utilizing bio-inspired steganography. We first construct bionic sonar signal waveforms which are very close to the true whale whistle, and then embed these constructed bionic sonar signal waveforms into the true whale call trains to hide the real sonar signal waveforms. According to the time-frequency structure of the true whale whistle, a bionic sonar signal model is established to generate the proposed sonar signal waveforms. A single sonar signal is used to measure the range of the target and a combination of two sonar signals is utilized for measuring its speed. A high-performance range and speed measurement algorithm is deduced in detail. Based on the constructed signal waveforms and the characteristics of false killer whale call trains, a camouflage application strategy is designed to improve the camouflage ability of the sonar signal sequence. Finally, simulation results are provided to verify the performance of the proposed method.
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.
The covertness of an active sonar is a very important issue and the sonar signal waveform design was studied to improve covertness of the system. Click is one of the most common cetacean sounds, and ...has excellent detection performance. Meanwhile, existing interception systems normally classify cetacean sounds as ocean noise and filter them out. Based on this, a bio-inspired covert active sonar detection method is proposed by encoding the clicks produced by sperm whales. First, based on the frequency distribution characteristic of sperm whale clicks and frequency-hopping sequence, a set of bionic sub-pulses are constructed by encoding the centroid frequency of the real clicks. Based on the inter-click interval characteristic of the clicks and time-hopping sequence, the bionic sonar signal is constructed by combining the constructed bionic sub-pulses and encoding the inter-pulse interval between two adjacent pulses. Then, a single bionic sonar signal is utilized to estimate the speed and range of the target. Based on the characteristics of sperm whale click trains, a camouflage application method is designed to improve the camouflage ability of the bionic sonar signal. Finally, experiment results show that when the target speed is 5m/s and 10m/s and the signal to noise ratio varies from −20dB to 20dB, the root-mean-square error of estimated speed and range are lower than 0.1m/s and 0.1m, respectively. In addition, the constructed bionic sub-pulses have a 99% probability of being classified into sperm whale clicks. Experimental results demonstrate the effectiveness and covertness of the proposed method.
A new algorithm called the Mondrian detector has been developed for object detection in high-frequency synthetic aperture sonar (SAS) imagery. If a second (low) frequency-band image is available, the ...algorithm can seamlessly exploit the additional information via an auxiliary prescreener test. This flexible single-band and multiband functionality fills an important capability gap. The algorithm's overall prescreener component limits the number of potential alarms. The main module of the method then searches for areas that pass a subset of pixel-intensity tests. A new set of reliable classification features has also been developed in the process. The overall framework has been kept uncomplicated intentionally in order to facilitate performance estimation, to avoid requiring dedicated training data, and to permit delayed real-time detection at sea on an autonomous underwater vehicle. The promise of the new algorithm is demonstrated on six substantial data sets of real SAS imagery collected at various geographical sites that collectively exhibit a wide range of diverse seafloor characteristics. The results show that-as with Mondrian's art-simplicity can be powerful.
Forward-looking sonar is one of the essential imaging equipment used in exploring underwater targets. However, it is always challenging to detect targets from sonar images considering the complex ...environment. This article represents an automatic underwater target detection method using clustering, segmentation, and feature discrimination. First, 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. Second, 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 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 popular deep learning approaches at present.