Real-time continuous active sonar processing

Canepa, Gaetano; Micheli, Michele; Morlando, Luca; Murphy, Stefan M.

doi:10.1109/oceans-genova.2015.7271658

Cited by 12 publications

(7 citation statements)

References 2 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, for long-range sensing using medium acoustic frequencies, the Doppler shift is relatively small, and a long, continuous wave (CW) probing pulse is required in order to detect the target's Doppler shift [25, ch.19]. A relatively new approach to overcome the slow update-rate problem is the concept of continuously active SONAR (CAS) [26], [27], [28], which applies processing on different frequency sub-bands of a long, wide-band acoustic transmission, thus achieving more detection opportunities, compared to the conventional pulse active SONAR.…”

Section: Related Workmentioning

confidence: 99%

Detecting Submerged Objects Using Active Acoustics and Deep Neural Networks: A Test Case for Pelagic Fish

Testolin

Kipnis

Diamant

2022

IEEE Trans. on Mobile Comput.

View full text Add to dashboard Cite

The accurate detection and quantification of submerged targets has been recognized as a key challenge in marine exploration, one that traditional census approaches cannot handle efficiently. Here we present a deep learning approach to detect the pattern of a moving fish from the reflections of an active acoustic emitter. To allow for real-time detection, we use a convolutional neural network, which provides the simultaneous labeling of a large buffer of signal samples. This allows to capture the structure of the reflecting signal from the moving target and to separate it from clutter reflections. We evaluate system performance both on synthetic (simulated) data, as well as on real data recorded over 50 sea experiments in a variety of sea conditions. When tested on real signals, the network trained on simulated patterns showed non-trivial detection capabilities, suggesting that transfer learning can be a viable approach in these scenarios, where tagged data is often lacking. However, training the network directly on the real reflections with data augmentation techniques allowed to reach a more favorable precision-recall trade-off, approaching an ideal detection bound. We also evaluate an alternative model based on recurrent neural networks which, despite exhibiting slightly inferior performance, could be applied in scenarios requiring on-line processing of the reflection sequence.

show abstract

Section: Related Workmentioning

confidence: 99%

Detecting Submerged Objects Using Active Acoustics and Deep Neural Networks: A Test Case for Pelagic Fish

Testolin

Kipnis

Diamant

2022

IEEE Trans. on Mobile Comput.

View full text Add to dashboard Cite

show abstract

“…7. It includes signal conditioning, possible sub‐band splitting, beamforming, matched filter (in the case of broadband waveform), normalisation, clustering, and finally contact detection (examples can be found in [147, 172, 173]).…”

Section: Acoustic Signal Processingmentioning

confidence: 99%

“…The matched filter output is normalised to reduce diffuse reverberation level and calculate the SNR per point. After the clustering of the measurements, an SNR threshold is applied to the normalised data to form contacts (also called detections) [147]…”

Section: Acoustic Signal Processingmentioning

confidence: 99%

“…One common requirement imposed by passive acoustic monitoring is persistence, which in general means lowpower systems; this may imply strong limitations on the measurement hardware design (number of array channels, computational power), hence on the system capabilities. Another main limitation in the use of small mobile robots for passive [147] surveillance is on the maximum physical aperture of the sensor array, if the targets are very low-frequency noise sources.…”

Section: Passive Sonar Signal Processingmentioning

confidence: 99%

See 1 more Smart Citation

Cooperative robotic networks for underwater surveillance: an overview

Ferri

Munafó

Teseï

et al. 2017

IET Radar Sonar & Navi

136

View full text Add to dashboard Cite

Underwater surveillance has traditionally been carried out by means of surface and undersea manned vessels equipped with advanced sensor systems. This approach is often costly and manpower intensive. Marine robotics is an emerging technological area that enables the development of advanced networks for underwater surveillance applications. In contrast with the use of standard assets, these advanced networks are typically composed of small, low-power, and possibly mobile robots, which have limited endurance, processing and wireless communication capabilities. When deployed in a region of interest, these robots can cooperatively form an intelligent network achieving high performance with significant features of scalability, adaptability, robustness, persistence and reliability. Such networks of robots can be the enabling technology for a wide range of applications in the maritime domain. However, they also introduce new challenges for underwater distributed sensing, data processing and analysis, autonomy and communications. The main thrust of this study is to review the underwater surveillance scenario within a framework of four research areas: (i) underwater robotics, (ii) acoustic signal processing, (iii) tracking and distributed information fusion, and (iv) underwater communications networks. Progress in each of these areas as well as future challenges is presented.

show abstract

“…CASs are beneficial for tracking moving targets because CASs provide near continuous updates, high precision, broadband reception, a high signal-to-noise ratio (SNR), and the capability to detect multiple targets [4][5][6][7][8][9]. These characteristics are based on the CW's signal structure, particularly in the difference of the duty cycle between a CW and a pulsed waveform.…”

Section: Introductionmentioning

confidence: 99%