The student poster competition at OCEANS’14 MTS/IEEE St Johns was held at the St Johns Convention Center September 14–19. The following students presented posters:
Loic Bernicola, Telecom Bretagne
Ahmed Nait Chabane, ENSTA Bretagne
Brian Claus, Memorial University of Newfoundland
Antonella Colucci, Univeristy of Exeter
André Figueiredo, University of Porto
David Illig, Clarkson University
Violeta Skoro Kaskarovska, Florida Atlantic University
Jie Li, University of Michigan
Hannan Lohrasbipeydeh, University of Victoria
José Melo, University of Porto
Tushar Mohan, Singapore University of Technology and Design
Jean-Christophe Morgére, Telecom Bretagne
Parvathy K.G., National Institute of Technology Karnataka
Himansu Pradhan, Indian Institute of Technology Delhi
Juan Ramirez, Florida Atlantic University
Ignacio Sánchez-Gendriz, University of Sao Paulo
Ken Sooknanan, Trinity College Dublin
Anthony Sylvester, University of Hawaii at Monoa
Jeffrey Walls, University of Michigan
Bo Li, Zhejiang University
The winners of the competition were:
Loic Bernicola, Telecom Bretagne, A Hybrid Registration Approach Combining SLAM and Elastic Matching for Automatic Side-Scan Sonar Mosaic
Abstract—This paper introduces a hybrid registration approach to build mosaics from side-scan sonar images. Due to specific acquisition procedure during surveys, standard SLAM techniques may not be robust enough to globally take into account a complete survey and correct sensor trajectories in order to properly georeference every pixel from all these images. iSAM algorithm has been fed with real side-scan images and shows interesting capabilities to produce corrected sensor trajectories allowing relevant coarse image registration, based on landmarks extraction and pairing. These trajectories will then guide a blockmatching procedure that will refine these trajectories by finely matching only sonar images relevant areas.
Ahmed Nait Chabane, ENSTA Bretagne, Unsupervised Knowledge Discovery of Seabed Types using Competitive Neural Network: Application to Sidescan Sonar Images
Abstract—The conventional approaches for habitats mapping based on supervised algorithms need a seabed ground truth classes to know the entire seabed types before the training phase. These approaches give satisfying results only when a comprehensive training set is available. If the training set lacks a particular kind of seabed, it will be unknown for the classifier and the classification will be reduced to the closest known sediment class. In addition, it is not always feasible to have a ground truth samples and generally costs are very important. This is what, automated sonar systems classification are becoming widely used. This paper is concerned with automated discovery of seabed types in sonar images. A novel unsupervised approach based on competitive artificial neural network (CANN) for sidescan sonar images segmentation is proposed. The main idea is to create an unsupervised color table which allows linking between the class color and the physical nature of the seabed. This process is based on these steps. The first one consists on texture features extraction from sonar images. Secondly, Self-Organizing features maps (SOFM) algorithm is used to project the vector features on two dimensional map. Then principal component analysis (PCA) is applied to reduce the dimensionality of the result of SOFM map to only three components. The three axes obtained by PCA process will be present the RGB color table. The final result of the color table can be used for supervised or unsupervised classification of sidescan sonar images.
Brian Claus, Memorial University of Newfoundland, Towards Navigation of Underwater Gliders in Seasonal Sea Ice
Abstract—The suitability of the available navigational aids for underwater gliders for year round use in waters which experience seasonal sea ice is evaluated and a path towards an operational system on the Labrador Shelf is presented. The extent of ice coverage is generally found to be limited to the shelf areas and with a duration of up to 20 weeks. For a desired navigational accuracy of 100 meters over a potential trackline in from the shelf break and back out again, around 400 kilometers, a series of low frequency sound sources or geophysical navigational methods are proposed. Acoustic methods require more maintenance and are more prone to loss, while geophysical methods require additional evaluation in the operational region and potential digital elevation model refinement. A three phase strategy is proposed to enable under ice observations. The first phase involves operating the gliders in the ice free season over the proposed track-lines. This data collection phase would allow the evaluation of the available methods and build confidence for later under ice operations. The second phase involves the refinement of the available DEMs both bathymetric and magnetic to the degree that successful navigation by geophysical methods is achieved during the ice free season. Upon the success of the vehicles navigation without surface access during the ice free season, the third phase would commence, that of under ice observations.
Antonella Colucci, Univeristy of Exeter, Investigating the Interaction of Waves and Currents from ADCP Field Data
Abstract—This paper characterises wave and current fields at two different marine energy site with investigation of WCIs (wave current interactions) and turbulence. The features of a 5-beam ADCP are utilised whereby it is found to accurately measure wave spectra up to high frequencies as well as providing directional wave data. Deployments were conducted in 2010 and 2011 using a Teledyne RDI 5-beam ADCP (Acoustic Doppler Current Profiler) recording waves and currents at 2Hz at the North Cornwall Wave Hub site in the UK. While the one conventional four beam 600 kHz unit ADCP was deployment at The Pentland Firth off the North coast of Scotland in March 2013. Current velocity off the North Cornwall coast is influenced principally by the semi-diurnal variation in tidal flow and by the regular cycle of spring and neap tides, with a maximum of 1 m/s Current Magnitude. The interaction between waves and currents has been studied and the important of the direction angle between wave of current has been under light in order to find appreciate interaction in the site. In the second site up of Scotland there are completely different sea state with the current magnitude up of 5 m/s where it is shows clearly the Doppler Shift effect and the wave direction influenced by the current. Most places in the world we can safely assume that the water velocity we measure is predominantly a superposition of waves and currents, but in the Penland Firth, The new assumption is that the measured velocity is really a superposition of waves, currents and turbulence, it is fundamental to investigate for the design of Marine Energy Converters (MECs).
André Figueiredo, University of Porto, Tracking Of An Underwater Visual Target With An Autonomous Surface Vehicle
Abstract—This paper presents the development of a first approach to a vision-based target detection. The ultimate objective of this work is to position an autonomous surface vehicle relative to a target. Experiments in a controlled indoor environment were conducted to test the developed system. The experimental results are analyzed and show that the tracking performances achieve errors in the order of a few centimetres.
David Illig, Clarkson University, Statistical Backscatter Suppression Technique For A Wideband Hybrid Lidar-Radar Ranging System
Abstract—A new backscatter suppression technique is applied to a wideband modulation scheme to enhance optical ranging in underwater environments. The statistical digital signal processing (DSP) approach of blind signal separation (BSS) [1-2] is applied to a frequency domain reflectometry (FDR) [3-4] ranging system. Applying BSS to the FDR system allows the backscatter return to be dynamically measured and cancelled out before computing range. Results from simulations and laboratory experiments are presented to demonstrate the combined FDR/BSS approach.
Violeta Skoro Kaskarovska, Florida Atlantic University, Performance Analysis of the Single Input Multiple Output Acoustic System for High Frequency Shallow Water Communications
Abstract—A Single Input Multiple Output (SIMO) acoustic communication system has been developed to improve high data rate communications at short range in the shallow water acoustic channel. The proposed communication system operates at very high frequency and combines an adaptive spatial diversity and parallel Decision Feedback Equalizer (DFE). Experimental results using selective combining with three, four and five receivers demonstrate ability to improve the quality of the acoustic channel in terms of Bit Error Rate (BER) and Signal-to- Noise-and-Interference Ratio (SNIR). The work presented in this article focuses on statistical evaluation of the proposed SIMO system using simulated data. For that purpose, a simulation channel model is developed based on experimental channel model and Rician fading channel. Adaptive multilevel combining is performed on simulated messages with high individual BER before SIMO processing. The simulation results achieved with selective diversity and SIMO system with three, four and five receivers show that the BER and SNIR of the combined message improve dramatically, compared to the BER and SNIR of the individual message and demonstrate that the proposed SIMO system can significantly improve the reliability of the shallow water acoustic channel.
Jie Li, University of Michigan, Multi-altitude Multi-sensor Fusion Framework for AUV Exploration and Survey
Abstract—In this paper, we propose a path planning framework for underwater exploration and rugosity estimation using Autonomous Underwater Vehicles (AUVs). Rugosity, a measure of variation in the height of a surface, is commonly used to characterize seafloor habitat. The goal of this work is to optimize the survey of an unknown area in order to efficiently estimate its rugosity. To this end, we propose a novel rugosity approximation on 3D voxel grids and a novel framework for using that approximation to adaptively plan AUV paths. The proposed method employs a heterogeneous set of sensors – multibeam sonar and stereo cameras – whose varied resolution and range make them complimentary for this task. For broad-scale exploration, sonar is used to produce a coarse sense of the area’s structure. Fine-scale exploration is completed using the stereo cameras to refine the high-resolution estimate of rugosity. Results display the simulation of two scenarios on real structural data gathered with an AUV and diver held sensor. The first scenario explores the situation where no broad-scale information is available and the robot must explore the terrain optically. The second simulates the two-pass case and demonstrates our method’s ability to achieve high accuracy rugosity estimation faster than other survey planning approaches.
Hannan Lohrasbipeydeh, University of Victoria, Passive Energy Based Acoustic Signal Analysis for Diver Detection
Abstract—A passive energy based analysis of the acoustic signal from a diver is presented. The frequency components of the signal during inhaling and exhaling are investigated, as well as the periodicity of the signal features. The effect of different regulators on the power spectral density of the signals is evaluated. The energy based detection problem is formulated using the timefrequency Teager Energy Operator (TEO) as it can be used to develop an adaptive detector. To evaluate the proposed approach, experimental data is employed which was recorded in the open ocean at different ranges and depths. This data was obtained using accurate hydrophones provided by Ocean Network Canada (ONC).
José Melo, University of Porto, A PHD Filter for Tracking Multiple AUVs
Abstract—In this paper we address the problem of tracking multiple AUVs using acoustic signals. Using For this challenging scenario, we propose to use a Probability Hypothesis Density Filter and present a suitable implementation of the Sequential Monte Carlo PHD filter. It will be demonstrated that a particle filter implementation of the aforementioned filter can be used to successfully track multiple AUVs, changing in number over time, using range measurements from the vehicles to a set of acoustic beacons. Simulation results will be presented that allow to evaluate the performance of the filter.
Tushar Mohan, Singapore University of Technology and Design, Modular Autonomous Robotic Fish
Abstract—the paper is subjected to a novel design of an autonomous biomimetic robotic fish. The novel approach culminates in the usage of cable mechanics to mimic aquatic life form in specificity the Koi fish. This paper presents a holistic study and design architecture of the underwater robot. The modularity and accessibility of the design set it apart from any previous bio-mimicry experimentation or productization. This is chiefly due to the use of a polycarbonate spine maneuvered by a single servo motor and a separate module for the head to contain the microprocessor controlling the sensory data. Reduction of complexity for the common user and number of day to day underwater applications without disrupting the ecosystem are few of the real-world outcomes with the implementation of this principle of robotic fish.
Jean-Christophe Morgére, Telecom Bretagne, Electronic Navigational Chart Generator for a Marine Mobile Augmented Reality System
Abstract—Augmented Reality devices are about to reach mainstream markets but applications have to meet user expectations in terms of usage and ergonomics. In this paper, we present a real-life outdoor Marine Augmented Reality Navigational Assistance Application (MARNAA) that alleviates cognitive load issues (orientation between electronic navigational devices and bridge view) for vessels and recreational boats. First, we describe the current application and explain the requirements to draw relevant and meaningful objects. Secondly we present the 3D chart generator that extracts and provides the meaningful information to the application. Then, we detail our Marine Mobile Augmented Reality System (MMARS) and a generic architecture that can embeds MARNAA application. Finally, we present results and implementations.
Parvathy K.G., National Institute of Technology Karnataka, A Numerical Modeling Approach for Study of Mudbank Impact on Coastline
Abstract—Coastal zone is the triple interface of land, ocean and atmosphere. Any developmental activity along the coastal zone requires a clear understanding of the dynamic processes controlling its very existence. When most of the processes, which are common to all coastlines are quite well known, there are some localized, but important processes requiring further research for developmental planning. Mudbanks are such an inquisitive coastal phenomenon which occur only at a few locations in the nearshore waters of the world ocean. Mudbanks, its occurrence, nature, properties and characteristics are interesting subjects from engineering point of view. The objective of the study is to provide an insight of mudbank impact on coastal morphology through a numerical modeling approach. For a better understanding of the influence of mudbanks on coastal morphology, Munambam to Chettuwa sector of Thrissur coast, which is a part of Southwest coast of India, is considered. In the present study the description of coastline evolution due to impact of mudbank is calculated using LITLINE module of LITPACK software package. It is observed that the occurrence, non-occurrence and migration of mudbanks influence the coastal dynamics significantly along mudbank influenced coastal stretch of Kerala.
Himansu Pradhan, Indian Institute of Technology Delhi, Internal Wave Simulation For Different Angles and shapes of continental shelf
Abstract—The average slope of the continental shelf in the world ocean is 0.5° and its width varies considerably. This paper illustrates experimental studies describing the internal wave run-up on different gradients of continental shelf varying from 0.2° to 0.5°. MIT general circulation model is configured with a variable grid, tidal information in the momentum equations and background stratification of density as initial fields to simulate internal waves. The model simulated density and temperature time-series is subjected to Fast Fourier Transform to compute the energy spectra of internal waves. The results reveal that the peak of internal wave activity varies spatially for different angles of the continental shelf. The experiments are further continued for concave coastline geometry to look at the internal wave energy distribution over the shelf. The results show that in a concave coastline the energy is large compared to a straight coastline inferring convergence of internal wave energy.
Juan Ramirez, Florida Atlantic University, Adaptive Torque Control of In-Stream Hydrokinetic Turbines
Abstract—In-stream hydrokinetic turbines, devices that generate electrical power from moving currentswithout the use of dams, have the potential to significantly impact the electrical power portfolio of many countries. This work focuses on improving the power production of in-stream hydrokinetic turbines by implementing an adaptive controller that regulates shaft torque. This controller is evaluated using a numerical rotor simulation that utilizes a blade element momentum approach with a dynamic wake in-flow model. The conducted numerical simulations demonstrate that adaptive control can effectively converge a sub-optimal controller gain to an optimal one that maximizes shaft power using only measured RPM and flow speed values.
Ignacio Sánchez-Gendriz, University of Sao Paulo, Harbor Sound Level Estimation For Assessment Of Underwater Ship Noise
Abstract—It is well known that rising anthropogenic underwater noise has chronic impacts on marine fauna. Since the contribution of navigation is particularly important to this kind of noise, areas near big harbors can be critically affected. One case of great environmental concernis the Port of Santos, the largest harbor in Latin America that, additionally, is expected to be expanded. The port is located on the brazilian southeast coast, a region with rich marine biodiversity including marine mammals, close to two marine conservation areas. Although evaluating impacts of ship noise in the area is a true concern, there is little information available on the topic nowadays. In this context, the present work aims to assess the magnitude of ship noise in a delimited geographical area around Santos harbor. For this purpose, Sound Pressure Level (SPL), Sound Exposure Level (SEL), and Power Spectral Density (PSD) were calculated. The SPL and SEL were similar to levels reported from other studied regions, reaching values that could interfere with some marine species. Results show that vessels noise is a major contributor to the ambient noise in the area assessed.
Ken Sooknanan, Trinity College Dublin, Mosaics For Nephrops Detection in Underwater Survey Videos
Abstract—Harvesting the commercially significant lobster, Nephrops norvegicus, is a multimillion dollar industry in Europe. Stock assessment is essential for maintaining this activity but it is conducted by manually inspecting hours of underwater surveillance videos. To improve this tedious process, we propose an automated procedure. This procedure uses mosaics for detecting the Nephrops, which improves visibility and reduces the tedious video inspection process to the browsing of a single image. In addition to this novel application approach, key contributions are made for handling the difficult lighting conditions in these kinds of videos. Mosaics are built using 1-10 minutes of footage and candidate Nephrops regions are selected using image segmentation based on local image contrast and colour features. A K-Nearest Neighbour classifier is then used to select the respective Nephrops from these candidate regions. Our final decision accuracy at 87.5% recall and precision shows a corresponding 31.5% and 79.4% improvement compared with previous work.
Anthony Sylvester, University of Hawaii at Monoa, Variable Buoyancy Control for a Bottom Skimming Autonomous Underwater Vehicle
Abstract—Two feedback controllers are presented that utilize data averaging and model-based estimation to offset the effects of sensor noise and achieve precise control of an autonomous underwater vehicle (AUV) variable buoyancy system (VBS). Operation of the bottom skimming AUV requires a constant reaction force between the seabed and the vehicle. While performing a mission, variable seafloor topography and a changing payload weight requires the use of a VBS to maintain the reaction force. Two traits of the VBS system that make this a challenging problem are the presence of sensor noise and fast on/off actuation relative to the sensor update rate. It was discovered that both controllers function under these conditions but the model-based controller provides more precise control of the system. This paper presents a comparison between these two control algorithms based on both simulation results and field experiments in a coastal environment.
Jeffrey Walls, University of Michigan, Toward Informative Planning for Cooperative Underwater Localization
Abstract—This paper reports on an algorithm for planning a practical trajectory for a surface vehicle that provides range measurements to an autonomous underwater vehicle (AUV). We consider server-client cooperative localization in which a server vehicle provides relative range constraints to minimize the uncertainty of a client vehicle. Our approach assumes the nominal client mission plan is available and draws potential server trajectories from a set of parameterized trajectory classes. We provide a comparative evaluation over several simulations, for both a single client and multiple clients, demonstrating that our algorithm computes operationally practical server paths and performs well relative to existing planning frameworks.
Bo Li, Zhejiang University, Simulation and Preliminary Experimental Results on S-surface Control of an Autonomous Underwater Vehicle Based on MOOS-IvP
Abstract—Classic PID algorithm is sensitive to the net buoyancy in the underactuated Autonomous Underwater Vehicle (AUV) control, and can be particularly problematic in the environment where the density of the water changes. S-surface control is based on the theory of fuzzy control and the structure of PD control, which shall be able to handle the above case well. It also simplifies the design of the controller and can improve the performance of non-linear control. In this paper, we test a number of cases by simulations of an underactuated AUV, especially investigating the effects of buoyancy of the AUV and the ocean current on the control performance. Further, we implement the S-surface control algorithm on a small AUV platform using MOOS-IvP, and present a preliminary experiment in a lake. The results show that the S-surface control has better performance compared to the PID in the case that the net buoyancy changes.