Underwater object detection and temporal signal detection in turbid water using 3D-integral imaging and deep learning.

Underwater scattering caused by suspended particles in the water severely degrades signal detection performance and poses significant challenges to the problem of object detection. This paper introduces an integrated dual-function deep learning-based underwater object detection and classification and temporal signal detection algorithm using three-dimensional (3D) integral imaging (InIm) under degraded conditions. The proposed system is an efficient object classification and temporal signal detection system for degraded environments such as turbidity and partial occlusion and also provides the object range in the scene. A camera array captures the underwater objects in the scene and the temporally encoded binary signals transmitted for the purpose of communication. The network is trained using a clear underwater scene without occlusion, whereas test data is collected in turbid water with partial occlusion. Reconstructed 3D data is the input to a You Look Only Once (YOLOv4) neural network for object detection and a convolutional neural network-based bidirectional long short-term memory network (CNN-BiLSTM) is used for temporal optical signal detection. Finally, the transmitted signal is decoded. In our experiments, 3D InIm provides better image reconstruction in a degraded environment over 2D sensing-based methods. Also, reconstructed 3D images segment out the object of interest from occlusions and background which improves the detection accuracy of the network with 3D InIm. To the best of our knowledge, this is the first report that combines deep learning with 3D InIm for simultaneous and integrated underwater object detection and optical signal detection in degraded environments.

Linear optoacoustic underwater communication.

The linear mechanism for optical-to-acoustic energy conversion is explored for optoacoustic communication from an in-air platform or surface vessel to a submerged vessel such as a submarine or unmanned undersea vehicle. The communication range that can be achieved is addressed. A number of conventional signals used in underwater acoustic telemetry applications are shown to be capable of being generated experimentally through the linear optoacoustic regime conversion process. These results are in agreement with simulation based on current theoretical models. A number of practical issues concerning linear optoacoustic communication are addressed that lead to a formulation of a linear-regime optoacoustic communication scheme. The use of oblique laser beam incidence at the air-water interface to obtain considerable in-air range from the laser source to the in-water receiver is addressed. Also, the effect of oblique incidence on in-water range is examined. Next, the optimum and suboptimum linear optoacoustic sound-generation techniques for selecting the optical wavelength and signaling frequency for optimizing in-water range are addressed and discussed. Optoacoustic communication techniques employing M-ary frequency shift keying and multifrequency shift keying are then compared with regard to communication parameters such as bandwidth, data rate, range coverage, and number of lasers employed.

Experimental demonstration of multiple pulse nonlinear optoacoustic signal generation and control.

Generating underwater acoustic signals from a remote, aerial location by use of a high-energy pulsed infrared laser has been demonstrated. The laser beam is directed from the air and focused onto the water surface, where the optical energy was converted into a propagating acoustic wave. Sound pressure levels of 185 dB re microPa (decibel re microPa) were consistently recorded under freshwater laboratory conditions at laser-pulse repetition rates of up to 1000 pulses/s. The nonlinear optoacoustic transmission concept is outlined, and the experimental results from investigation of the time-domain and frequency-domain characteristics of the generated underwater sound are provided. A high repetition rate, high-energy per pulse laser was used in this test under freshwater laboratory conditions. A means of deterministically controlling the spectrum of the underwater acoustic signal was investigated and demonstrated by varying the laser-pulse repetition rate.

Experimental demonstration of remote, passive acousto-optic sensing.

Passively detecting underwater sound from the air can allow aircraft and surface vessels to monitor the underwater acoustic environment. Experimental research into an optical hydrophone is being conducted for remote, aerial detection of underwater sound. A laser beam is directed onto the water surface to measure the velocity of the vibrations occurring as the underwater acoustic signal reaches the water surface. The acoustically generated surface vibrations modulate the phase of the laser beam. Sound detection occurs when the laser is reflected back towards the sensor. Therefore, laser alignment on the specularly reflecting water surface is critical. As the water surface moves, the laser beam is reflected away from the photodetector and no signal is obtained. One option to mitigate this problem is to continually steer the laser onto a spot on the water surface that provides a direct back-reflection. Results are presented from a laboratory test that investigates the feasibility of the acousto-optic sensor detection on hydrostatic and hydrodynamic surfaces using a laser Doppler vibrometer in combination with a laser-based, surface normal glint tracker for remotely detecting underwater sound. This paper outlines the acousto-optic sensor and tracker concepts and presents experimental results comparing sensor operation under various sea surface conditions.