OCEANS'17 Anchorage

This 41th Student Poster Program of the OCEANS Conferences was held at OCEANS ’17 MTS/IEEE Anchorage, at the Dena’Ina Convention Center, from September 18 to September 21. The program was organized by Liesl Hotaling (MTS SPC Chair) as local coordinator and Philippe Courmontagne, SPC Chair, from IEEE OES. The program was funded by grants from the Office of Naval Research. For this edition, 20 student abstracts were selected for this contest, not without difficulty given the high quality of the received abstracts. Students came from Brazil, Canada, China, France, Italy, Japan, Korea, Portugal, United Kingdom and the United States.
     The posters were on display in the Exhibition Hall. As for the previous Student Poster Competitions, outstanding posters describing the work of the students were presented and were particularly appreciated by the attendees of the conference. Moreover, the student participants greatly appreciated the opportunity to display, exchange and describe their research work to the community. The posters were judged by a team organized by IEEE OES and MTS. The roster of students and their schools are (in order of appearance of the Program Book):


  • Takayuki Nishimura, University of Tokyo, Japan
  • Renzheng Che, Ocean University of China
  • Arnold Kalmbach, McGill University, Canada
  • Yiheng Wang, Missouri University of Science and Technology
  • Jialei Zhang, Huazhong University of Science & Technology, China
  • Taeyun Kim, KAIST, Korea
  • Zhuoyuan Song, University of Florida
  • Thanh Huy Nguyen, IMT Atlantique Bretagne-Pays de la Loire, France
  • Seonghun Hong, KAIST, Korea
  • Gonçalo Cruz, Instituto Superior Técnico, Portugal
  • Byeongjin Kim, POSTECH, Korea
  • Kaitlyn Morgan, Clemson University
  • Cassidy Gonzalez-Morabito, Rutgers University
  • Adham Sabra, Robert Gordon University, United Kingdom
  • Jonathan Soli, Duke University
  • Muhammad Fahad, Stevens Institute of Technology
  • Bruno Floriani, PE-CAPES, Federal University of Santa Catarina, Brazil
  • Andrea Petroni, “La Sapienza” University of Rome, Italy
  • John McKay, Pennsylvania State University
  • Eduardo Iscar, University of Michigan
Liesl Hotaling

     The judging was completed by noon on Thursday and the prizes were awarded in the Exhibition Hall. The ceremony began with some few words from Liesl Hotaling, recalling the history of this Student Poster Competition, initiated and championed by Norman Miller, who passed away in July 2015.
     Then, Philippe Courmontagne called all of the students on stage and presented each student with a certificate for their participation in the program. Christian de Moustier, IEEE OES President, and Donna Kocak, MTS President, were called up to present the awards. The third prize was awarded to Zhuoyuan Song, from the United States, the second prize to Arnold Kalmbach, from the United States and the first prize, the “Norman Miller’s Prize” to Kaitlyn Morgan, for her poster entitled “Higher Order Bessel Beams Integrated in Time (HOBBIT) for Free Space Underwater Sensing and Communication”. As with previous years, monetary prizes were awarded for the posters collectively ranked 1st, 2nd and 3rd place by the judges ($3000, $2000 and $1000 respectively). The audience gave the students a big hand following the awards presentations. The session ended with a photograph session.
     The roster of students and their poster titles are given below with an abstract of their paper.


The awards ceremony.

Takayuki Nishimura, University of Tokyo
Toward automatic detection of deep sea top predators by visible light
     In this paper, we propose an automatic shark detection method in deep sea by light section method, in order to realize shark observation with much less energy. This method detects sharks on the basis of cross-sectional shape obtained by light section method. To analyze the performance of the proposed method, experiments at aquariums was conducted. It was observed that detection accuracy became high by setting thresholds considering physical characteristics of shark. Future work includes additional sampling and sea trial.

Renzheng Che, Ocean University of China
Online Fish Tracking with Portable Smart Device for Ocean Observatory Network
     Nowadays, ocean observatory networks, which gather and provide multidisciplinary, long-term, 3D continuous marine observations at multiple temporal spatial scales, play a more and more important role in ocean investigations. In this paper, we try to develop a portable smart device with online fish detection and tracking strategies by ARM7 microprocessor for ocean observatory networks, combining the deformable fish body description with compressive sensing together. A fish detection system is introduced to represent highly variable objects using mixtures of multiscale deformable fish body description which include three parts: (1) root filter, which describes the global contour feature of underwater fish. (2) fish body part filter, which capture finer resolution features of underwater fish. (3) spatial model, which describes the relationship between the fish body part filters and the root filter. And then, we apply a simple yet effective, real-time and robust of fast compressive tracking algorithm with an appearance model to focus on the motion trajectories of underwater fish. It has been shown in the simulation experiments that the developed scheme of this paper achieves consistent performance improvements on online fish tracking for ocean observatory network.

Arnold Kalmbach, McGill University
Learning Seasonal Phytoplankton Communities with Topic Models
     In this work we develop and demonstrate a probabilistic generative model for phytoplankton communities. The proposed model takes counts of a set of phytoplankton taxa in a timeseries as its training data, and models communities by learning sparse co-occurrence structure between the taxa. Our model is probabilistic, where communities are represented by probability distributions over the species, and each time-step is represented by a probability distribution over the communities. The proposed approach uses a non-parametric, spatiotemporal topic model to encourage the communities to form an interpretable representation of the data, without making strong assumptions about the communities. We demonstrate the quality and interpretability of our method by its ability to improve performance of a simplistic regression model. We show that simple linear regression is sufficient to predict the community distribution learned by our method, and therefore the taxon distributions, from a set of naively chosen environment variables. In contrast, a similar regression model is insufficient to predict the taxon distributions directly or through PCA with the same level of accuracy.

Yiheng Wang, Missouri University of Science and Technology
Dubins Curves for 3D Multi-Vehicle Path Planning Using Spline Interpolation
     This paper proposes a spline interpolation method to map 2D Dubins paths into 3D for multiple Autonomous Underwater Vehicles (AUVs) to visit multiple targets. The multitarget assignment and path planning problem is first modeled as a multiple Traveling Salesmen Problem (mTSP) and a three-step algorithm is used to solve the NP-hard integer programming problem with affordable computational complexity. Step 1 uses the Genetic Algorithm (GA) with the 3D Euclidean distances as the fitness function to assign multiple tasks to multiple AUVs; Step 2 designs the 2D Dubins paths for each AUV target sequence; and Step 3 maps the 2D Dubins paths to 3D paths via cubic spline interpolation. Finally, potential collision is detected among the resulting paths of the multiple AUVs and paths are discarded if any AUV is within the close vicinity of another AUV at any time. The simulation results show that the proposed algorithm yields shortest Dubnis paths most of the time and guarantees the G1 continuity. The probability of collision is very small if the multiple AUVs start the mission at the geometric center of the multiple targets.

Jialei Zhang, Huazhong University of Science & Technology
Complete Coverage Tracking and Inspection for Sloping Dam Wall by Remotely Operated Vehicles
     This paper focuses on the inspection for sloping dam wall by a remotely operated vehicle (ROV). First, a dedicated ROV for dam inspection is described. Second, a fixed-distance tracking control strategy is proposed to guarantee the fixed camera-to-subject distance in sloping dam inspection, which is instrumental in sonar image fusion for the large-scale dam wall with a certain inclination. In addition, the ROV can periodically surface up and correct its position by using the onboard GPS, such that the positioning accuracy of the ROV and its detected crack point can be improved. Third, in order to reject the unknown hydrodynamic coefficients, water intake and currents, a model-independent fuzzy PI controller is designed for robust tracking in dam inspection. Finally, some numerical and experiment results are given to illustrate its performance.

Taeyun Kim, KAIST
A Novel Terrain Information Measure for Terrain-Referenced Underwater Navigation
     The use of elevation changes in an undulating terrain surface can be an effective alternative for vehicle navigation in GPS-denied underwater environments, since subsea terrain elevation data can be obtained using sonar systems. The performance of terrain-referenced navigation varies significantly depending on how informative a given terrain is, however it is not straightforward to quantify the amount of information that can be provided by the terrain and predict the navigation performance. This study proposes a new terrain information measure by analyzing and quantifying the amount of information in terms of terrain roughness and uniqueness. The expected information is evaluated in the spatial frequency domain via Fourier transforms to maximize the computational efficiency of the quantification. To demonstrate the performance and utility of the proposed ideas, terrain-referenced navigation simulation results are shown and discussed.

Zhuoyuan Song, University of Florida
Cooperative Mid-Depth Navigation Aided by Ocean Current Prediction
     A novel inertial navigation system is proposed for small autonomous underwater vehicles in long-duration, large-scale operations where frequent surfacing and consistent bottom-locking are not desirable. This strategy utilizes the dynamics of the background flow to significantly mitigate the dead-reckoning error of an inertial navigation system. This is achieved by comparing the local ambient flow velocity against the velocity field prediction pre-calculated through solving the background flow dynamics. The vehicle’s attitude and linear velocity are estimated along with the vehicle’s position in order to obtain an in-situ estimation of the absolute background flow velocity. Estimation errors of the vehicle’s location, attitude and linear velocity are mutually correlated through continuously incorporating relative flow velocity measurements into state estimation. The proposed navigation system is implemented as a marginalized sequential Monte Carlo estimator, where the vehicle’s position is estimated through samples, and the vehicle’s attitude and linear velocity are inferred by Gaussian filters. Opportunistic information fusion among neighboring vehicles are achieved using covariance intersection. The performance of the proposed navigation system is analyzed in simulation within a turbulent multi-gyre flow field.

Thanh Huy Nguyen, IMT Atlantique Bretagne-Pays de la Loire
Heterogeneous Data Registration for 3D Underwater Scene Reconstruction
     This paper addresses the heterogeneous data registration problem, which is one of the key features for any scene reconstruction and representation, especially for the underwater environment. In this study, we propose a registration method built around a 2D-to-3D feature-based approach that registers high-resolution side-scan sonar images with bathymetric data (topographic 3D point cloud) obtained by multibeam echosounder. This process enables us to achieve a global 3D mosaic of the studied underwater scene, which is informatively richer and more reliable than each individual dataset. Indeed, the interest of this data fusion representation is that it combines the benefits of using each sensor: bathymetric information provides the geometric structure of the sea-bottom, while sidescan sonar images contribute a complementary observation with better resolution of the sea-bottom reality (e.g. sedimentology, bottom-laying object, etc.)

Seonghun Hong, KAIST
Development of a Hover-Capable AUV System for Automated Visual Ship-Hull Inspection and Mapping
     Generally, underwater hull inspection have been conducted by human divers. It is an extremely dangerous task, and hence, can be a potential application for unmanned underwater vehicles. The operational safety and performance of in-water inspection can be significantly improved by introducing unmanned vehicle systems. This study addresses the development of an hover-capable autonomous underwater vehicle system and its operational algorithms for automated visual ship hull inspection with no (or minimum) human intervention. The feasibility and practical performance of the developed system and algorithms are demonstrated by conducting field experiments with a full-scale ship in a real sea environment.

Gonçalo Cruz, Instituto Superior Técnico
Evaluating Aerial Vessel Monitoring In Maritime Surveillance Scenarios Using Convolutional Neural Networks
     In this paper we present an autonomous detection approach for airborne surveillance in maritime scenarios. This approach is robust to sun glare, waves and scale variation. Additionally, we introduce a new metric to evaluate detection and tracking results that is more adequate for these scenarios. The proposed detection method is evaluated using videos from different monitoring missions and its results are compared with a state-of-the-art neural network. This comparison is done using a traditional and the proposed evaluation metric.

Byeongjin Kim, POSTECH (Pohang University of Science and Technology)
Imaging sonar based navigation method for backtracking of AUV
     We propose an imaging sonar-based backtracking method as a navigation strategy for an underwater investigation of AUVs. The purpose of backtracking method is to reduce a drift error caused by the inaccuracy of navigation sensors when an AUV returns to its previous position. The AUV divides the trajectory into several intervals and returns to the previous position while correcting small drift error for each interval. For the backtracking, we suggest a method obtaining terrain information using an imaging sonar. We create a 3D point cloud by scanning the seafloor. We can use the 3D point cloud to detect objects and to select natural landmarks. The selected natural landmark can be used as a reference in the backtracking process. To verify the feasibility of proposed methods, we conducted field experiments using a hovering-type AUV ‘Cyclops’.


Kaitlyn Morgan, Clemson University
Higher Order Bessel Beams Integrated in Time (HOBBIT) for Free Space Underwater Sensing and Communication
     This work presents the generation, dynamic phase modulation and phase detection of interfering composite optical vortices. This technique demonstrates a new method for encoding information onto an optical beam for potential use in underwater communication as well as sensing applications including object detection and channel characterization.


Cassidy Gonzalez-Morabito, Rutgers University
Flight Dynamics of Slocum Gliders in Intermediate-Water Hurricane
     The new and emerging use of Slocum gliders in oceanographic data collection has brought about a progressive new way to record in-situ oceanic data. This piece of technology has been used by the Rutgers University Department of Marine Science through over 400 deployments, including a deployment during Hurricane Sandy in late October 2012. Data being collected during this deployment was done by Slocum glider RU23, which recorded measurements such as backscatter, fluorescence, water velocity, conductivity, temperature and depth. The dynamics of water motion RU23 flew in caused an unforeseen additional dataset to be collected by the glider. Because RU23 flew during intermediate type ocean waves, the orbital motion of the water potentially caused fluctuations seen in the dive and climb profiles of RU23’s pressure record. Analysis can be done on the signal seen in the pressure record, by viewing its pressure fluctuation in the frequency spectrum. From this, wave period can be extracted. Accuracy of the glider’s extracted wave period will be done by comparing the glider wave period to recorded wave period done by nearby NDBC buoys. Also analyzed was the glider’s response to the turbulent water dynamics gauged by its performance in its attitude measurements.

Adham Sabra, Robert Gordon University
Dynamic Localization Plan for Underwater Mobile Sensor Nodes using Fuzzy Decision Support System
     Underwater mobile sensor node localization is a key enabling technology for several subsea missions. A novel scalable underwater localization scheme, called Best Suitable Localization Algorithm (BLSA), is proposed to dynamically fuse multiple position estimates of sensor nodes using fuzzy logic, aiming at improving localization accuracy and availability along the whole trajectory in missions. Numerical simulation has been conducted to demonstrate significant improvement in localization accuracy and availability by using the proposed fuzzy inference system. The proposed method provides a cost-effective localization system by harnessing all available localization methods on-board.

Jonathan Soli, Duke University
Time Domain Bandwidth Synthesis for Active Sonar
     This paper presents a technique for extending the bandwidth of chirped active sonar signals. Bandwidth synthesis (BWS) results in improved range resolution, increased signal-to noise ratio (SNR) and reduced fading due to delay-spread multipath. The approach demonstrated uses a low-order autoregressive (AR) process to model de-chirped segments of the received signal, which are then extrapolated using linear prediction (LP). Both simulated results and an initial sea trial result confirms BWS performance.

Muhammad Fahad, Stevens Institute of Technology
Evaluation of Ocean Plume Characteristics using Unmanned Surface Vessels
     We present results from a series of tracer dye experiments where ocean plume concentration is measured using unmanned surface vessels. The goal of the study is to characterize the fine-scale spatiotemporal plume structure which is used for developing and evaluating autonomous robotic sampling strategies. We present a description of the qualitative characteristics of experimentally generated marine plumes namely intermittency, sinuous structure, and the time varying near-source concentration profile. We also present a data reduction process and a set of summary statistics to describe the fine-scale plume structure as evidenced by the time series measurements. These summary statistics provide a comparison benchmark for the development of future plume simulation models that capture the fine-scale plume structure.

Bruno Floriani, Federal University of Santa Catarina
Model-Based Underwater Inspection via Viewpoint Planning using Octomap
     During the last years, the underwater inspection of industrial assets, archaeological shipwrecks or geological/ biological structures has become an area of interest. Nowadays, these inspections are done by divers or remotely operated vehicles (ROVs). This paper presents a preliminary solution for planning automatic underwater inspections for an autonomous underwater vehicle (AUV) assuming an a priori known environment. The proposed algorithm computes the viewpoints that an AUV has to reach in order to completely scan the region/structure of interest. Then, two optimization steps are taken into account: the number of viewpoints to cover the object and the order and path to reach them. Despite the algorithm uses a brute force approach, the utilization of octree based structures to represent the environment allows to obtain satisfactory results in a short time. The proposed method is evaluated in simulation using Gazebo and the Girona 500 AUV.

Andrea Petroni, “La Sapienza” University of Rome
On the MIMO Multipath Channels Spatial Correlation in Shallow Water Communications
     Acoustic communications in underwater environment are considerably influenced by the physical characteristics of the medium. This is even more true in the polar regions where the presence of icebergs and icy water layers represent an additional dependence to the signal propagation. Considering a MIMO scenario, we investigate how the channels spatial correlation is conditioned in the case of both frozen and fluid water surface. More, we evaluate the impact of the medium changes on the signal delay spread.

John McKay, Pennsylvania State University
Using Frame Theoretic Convolutional Gridding for Robust Synthetic Aperture Sonar Imaging
     Recent progress in synthetic aperture sonar (SAS) technology and processing has led to significant advances in underwater imaging, outperforming previously common approaches in both accuracy and efficiency. There are, however, inherent limitations to current SAS reconstruction methodology. In particular, popular and efficient Fourier domain SAS methods require a 2D interpolation which is often ill conditioned and inaccurate, inevitably reducing robustness with regard to speckle and inaccurate sound-speed estimation. To overcome these issues, we propose using the frame theoretic convolution gridding (FTCG) algorithm to handle the non-uniform Fourier data. FTCG extends upon non-uniform fast Fourier transform (NUFFT) algorithms by casting the NUFFT as an approximation problem given Fourier frame data. The FTCG has been show to yield improved accuracy at little more computational cost. Using simulated data, we outline how the FTCG can be used to enhance current SAS processing.

Eduardo Iscar, University of Michigan
Multi-view 3D Reconstruction in Underwater Environments: Evaluation and Benchmark
     Mapping of underwater environments is a critical task for a range of activities from monitoring coral reef habitats to surveying submerged archaeological sites. In recent years, optical reconstruction methods developed for terrestrial (in air) applications have increasingly been applied to the underwater environment by the marine science community. However, assumptions such as the brightness constancy constraint (BCC) and traditional camera pinhole models, which hold in air, do not apply in the underwater environment. There is a lack of literature about how the violation of these assumptions affects the accuracy of the reconstruction. The main contributions of this paper is to develop a quantitative evaluation of computer vision-based 3D reconstruction methods applied to the underwater domain. Comparisons are presented between low- and high-cost camera systems to quantify the tradeoff between equipment cost and performance.

A blind test? Not really that ... After many years and hundreds of students being chaired, as the SPC chair, new responsibilities come to me. So, it is just the time to give the mike to a new one:Prof. John Watson. I will still be there, no more as the SPC chair, but probably as a judge (I feel that I will regret having written this).
Keep in mind: Do not forget to join our huge family!