Paul J. Seekings

Visual and Passive Acoustic Marine Mammal Observations and High-Frequency Seismic Source Characteristics Recorded During a Seismic Survey

In this paper, we present marine mammal observation statistics, high-frequency seismic source characteristics, and example denoising of marine mammal acoustical recordings using data collected during the mitigation and monitoring program for a 3-D seismic survey by EnCana Corporation, Calgary, AB, Canada, in the Northwest Atlantic during 2003. Marine mammals were observed both visually and acoustically. No marine mammal incidents or adverse reactions were observed during the survey. Acoustical observations were made by the Sea map Passive Acoustic Cetacean Monitoring System (SPACMS), consisting of two hydrophones placed 50 m apart, towed ahead of and to one side of the seismic source. Visual and acoustical detections were uncorrelated, indicating the complementary nature of the two observational techniques. Visual detections were more common per hour of effort than acoustical detections. Acoustical detection rates showed no significant day-night difference. Marine mammals appeared to have avoided very close ranges (100 m) from the seismic array during seismic acquisition, but the overall number of marine mammals in the observable radius (1-2 km) did not change significantly when the seismic source was ldquoonrdquo compared to ldquooff.rdquo Marine mammals were observed in larger groups and appeared to have become less vocal when the seismic source was active. It should be noted however, that the results from this data gathering effort may be affected by potential sources of bias (such as the combination of data from toothed and baleen whales). Signal processing of seismic source signatures indicated some high-frequency energy content consistent with expectations from earlier work. This analysis confirmed that most of the seismic energy was concentrated at lower frequencies (500 Hz). No low-frequency comparisons with near-field data could be made due to the geometry of the SPACMS recording hydrophones and seismic source, which resulted in the Lloyds mirror effect obliterating low-frequency components in the SPACMS records. A wavelet-based denoising method was applied to improve the visibility of marine mammal vocalizations on a spectrogram display.

Classification of a large collection of whistles from Indo-Pacific humpback dolphins (Sousa chinensis)

The whistle repertoires of some dolphin species have been characterised and classified both qualitatively and quantitatively [1, 2, 3, 4, 5]; however, given the extensive range of the species relatively little is known about the acoustic repertoire of the Indo-Pacific humpback dolphin, Sousa chinensis [6, 7], particularly those in South-east Asia. Quantitative classification methods have been developed using sub-sets of S. Chinensis whistles [8, 9, 10] but research aimed at classifying a more complete whistle repertoire of S. Chinensis through quantitative classification methods has not been reported. To address this, whistles produced underwater by six Indo-Pacific humpback dolphins kept by Underwater World Singapore Pte. Ltd. at Dolphin Lagoon, Sentosa, Singapore were recorded. The acoustic recordings were de-noised and the whistles ‘traced’ from the spectrogram time/frequency representation. They were then classified using PCA and k-means clustering.

Classification of marine organisms in underwater images using CQ-HMAX biologically inspired color approach

In many coastal environments, particularly in tropical zones, coral reef ecosystems have exceptional biodiversity, contribute to coastal defense, provide unique and important habitats and valuable commercial resources. Assessment of environmental impacts on biodiversity in such areas are increasingly important to mitigate potential adverse effects on specific ecosystems. Visual classification of marine organisms is necessary for population estimates of individual species of corals or other benthic organisms. In this paper, we introduce a new image dataset of benthic organisms that are of different colors, shapes, scales, visibility and are taken from different viewpoints. We evaluate several different classification approaches on this dataset, and show that CQ-HMAX, our new biologically inspired approach to utilizing color information for object and scene recognition, that is inspired by the characteristics of color- and object-selective neurons in the high-level inferotemporal (IT) cortex of the primate visual system, results in better classification results in comparison with existing computational models such as support vectors machines, SIFT based approaches and the HMAX biologically inspired approach. We show that concatenating our model which encodes color information with the HMAX model which encodes grayscale shape information results in the highest classification accuracy.

Acoustics of shape recognition by a dolphin in a cross‐modal matching‐to‐sample paradigm

A pacific bottlenose dolphin was trained in a two‐alternative cross‐modal matching‐to‐sample paradigm. The animal was able to inspect complex PVC‐pipe shapes through echolocation or vision but never through both senses simultaneously. Acoustic data was collected through a 3‐channel high‐frequency recording system while the dolphin performed one of the following tasks: (1) match a complex shape with its sonar only (pure echoic matching), (2) match from vision to echolocation and (3) match from echolocation to vision. Simultaneously, synchronized in‐air and underwater video was recorded documenting the approach path of the dolphin to either the sample object or to the alternative objects. The collected data was analyzed for type of click signals used, the frequency range of the emitted clicks, number of clicks emitted before a successive match and variations of click type with different objects.

Denoising Dolphin Click Series in the Presence of Tonals, using Singular Spectrum Analysis and Higher Order Statistics

We examine the use of Singular Spectrum Analysis (SSA) technique as an alternative technique to using standard wavelet shrinkage schemes for the purpose of de-noising mixtures of tonals, transients and Gaussian noise. Wavelet schemes require a calculation of a threshold to determine which components are taken to be signal and noise. If the noise component is Gaussian, then threshold can be determined by using an appropriate estimator. However, in the presence of strong tonal content the Gaussian threshold estimators do not give optimal performance. One method is to iteratively shift the threshold until some performance criterion has been maximized. However this frequently leads to over de-noising this time series. Since the wavelet basis is chosen to best represent the signal of interest, over de-noising can cause artifacts to appear similar to the signal of interest. In most applications this can not be tolerated. SSA has advantages in that the basis of decomposition is derived from the time series itself. So-called Empirical Orthogonal Functions (EOFs) are derived from a lag matrix created from the time series. Singular Value Decomposition (SVD) is then used to decompose a time series into a number of time series components. In the case of signal separation or de-noising the time series components can be combined by using their statistical properties. We examine the use of higher order statistics, to group components into tonals, transient, and Gaussian noise. By using the properties of the kurtosis for these three types of signal, the grouping of components can be done in a more formal manner, than the thresholding technique found in wavelet schemes. The technique is demonstrated on test data consisting of dolphin clicks in the presence of tonal and Gaussian noise. Results are also shown for real data of a dolphin click series while echo-locating on a target. It is critical for future work that after de-noising, the shape of the dolphin clicks is preserved, and the recorded reflections from the target are adequately de-noised, without introducing artifacts which could be mistaken for reflections. We discuss the results of the SSA and evaluate its potential for de-noising applications.

Wavelet De-noising with Independent Component Analysis for Segmentation of Dolphin Whistles in a Noisy Underwater Environment

Bottlenose dolphins (Tursiops truncatus) are the most widely studied species of dolphin and are known to produce a complex mixture of different types of sounds. They are believed to communicate through frequency-modulated pure tones (whistles), and produce broadband clicks or click trains for echolocation while investigating their environment. They also produce a large range of other types of sounds variously described as barks, grunts, groans, etc. To further our aim of 2-way acoustically mediated communication with dolphins to study dolphin cognition, we need to separate Bottlenose dolphin whistles from noisy underwater recordings, which not only consist of whistles, but also broadband echolocation clicks, water splashes and other sources of ambient noise. Independent component analysis (ICA) has been successfully used for the separation of independent sound sources in many applications. In this paper we will discuss the use of ICA to separate dolphin whistles from other underwater sound sources.

The use of optical and sonar images in the human and dolphin brain for image classification

In this paper we propose a new biologically inspired model which simulates the visual pathways in the human brain used for classification of matching optical and sonar derived images. Marine mammals, such as dolphins, that live in waters with poor optical clarity and low light levels such as littoral zones, use a combination of optical vision and biosonar to navigate and hunt for prey. Given that dolphins have evolved a synergistic combination of optical visual input and acoustic/sonar input, the primary focus of this paper is on reaching a similar level of synergy for a diver or Autonomous Underwater Vehicle (AUV) platform equipped with a system to extend the range and resolution of vision in poor ambient visibility. We propose a biologically inspired model that combines and processes visual images acquired via optical and acoustic pathways and show that the combined model enhances the accuracy of automatic classification of target objects in underwater images.

Measuring the augmented sound localization ability of humans in the underwater environment

Humans are poor, if not incapable, at localizing sound underwater due to significant reduction in Inter-aural Temporal Differences (ITD) and Inter-aural Intensity Differences (IID) caused by reduced impedance mismatch and the higher sound speed in water. An improvement in sound localization underwater will significantly enhance divers safety, the way divers perceive and appreciate the underwater environments. A system that augments and enhances the sound localization ability of humans underwater was built for this purpose. The system extracts directional cues from high frequency acoustic component of the received signal and reintroduce the cues in audio band to the diver that wears the system. The novelty of this approach is that it does not need any explicit information on the signals in advance to localize them. The system passes almost the entire signal band to its user with minimum relative distortion except the directional cue ti re-introduced. It is then up to the user to perceive, detect, and localize the sound. In this paper, we present the setup and results from an experiment that measures the localization performance of divers using the system. The experiment setup consists of a source transmitter that was randomly positioned in a contiguous, one-meter radius, semi-circular frame, and a blindfolded subject that attempts to localize the acoustic source. Both the headings of the transmitter and subject were digitally recorded and compared to gauge the localization performance. Experiments have been carried out across different signal to noise ratio and across different frequencies above 20kHz. The result from the experiment shows that a diver using the system was able to localize a source to within ±15 degrees nearly 75% of the time. It is also observed that SNR does not significantly affect the localization performance within the range of SNR that we were testing. The subjects were able to localize acoustic source in a noisy marina environment with the system. The localization performance of the subjects seemed to improve as the subjects gained experience using the system over a few experiment sets. This suggests that the human brain adapts its perception ability and learns to use the new directional cues rather quickly.

Numerical Modeling of a Time Reversal Experiment in Shallow Singapore Waters

The time reversal mirror (TRM) technique is a very useful tool in many underwater applications. Its usefulness in reverberation rejection and underwater communications has already been established through experiments by the Marine Physical Laboratory. The design of a TRM experiment is specific to the location and environment where it is being conducted. This paper presents theoretical and numerical analysis of a time reversal experiment which will be conducted in very shallow water (15 -20 m depth) in Singapore waters. The objective of the numerical simulation was to arrive at the various design parameters for the experiment and thus to predict its performance. The main parameters under question were the optimum frequencies to be used and the focusing ranges to be investigated. Extensive measurements were carried out at the selected site to obtain information about the ambient noise, time evolving sound speed structure and also the sound velocity in the sea-bed. The bottom sound speed was computed from the bulk density and porosity of the core samples collected from various locations at the site using an empirical formula Direct measurements were also done to find out the propagation losses at three different frequencies (7.5, 10 and 12.5 kHz) and at three different depths (4, 8 and 12 m) over a 500 m range. A short description of the system hardware also is presented.

Perfomance Evaluation Of A Singal Crystal Hydrophone

Relaxor single crystals such as PZN-PT and PMN-PT exhibit superior electromechanical properties and are touted as the next-generation materials for future high performance piezo devices. This work describes a new hydrophone made of high-sensitivity PZN-PT single crystal d31 sensing elements. Three such elements are mounted in custom-made housing. A compartment is provided to house the three pre-amplifiers, one for each element. In the current prototype each element has its own input. This allows the directionality of each element to be measured and the possibility of locating the source direction from the combined signals of the three elements. A compact low noise high impedance voltage follower pre-amplifier has been designed which provides 26 dB of gain. The overall sensitivity was measured to be -169 dB re 1 V/muPa. Directionality has been measured to be approximately omni-directional within plusmn1 dB, up to a frequency of 7 kHz.