A SIFT-Like Feature Detector and Descriptor for Multibeam Sonar Imaging

Multibeam imaging sonar has become an increasingly important tool in the field of underwater object detection and description. In recent years, the scale-invariant feature transform (SIFT) algorithm has been widely adopted to obtain stable features of objects in sonar images but does not perform well on multibeam sonar images due to its sensitivity to speckle noise. In this paper, we introduce MBS-SIFT, a SIFT-like feature detector and descriptor for multibeam sonar images. This algorithm contains a feature detector followed by a local feature descriptor. A new gradient definition robust to speckle noise is presented to detect extrema in scale space, and then, interest points are filtered and located. It is also used to assign orientation and generate descriptors of interest points. Simulations and experiments demonstrate that the proposed method can capture features of underwater objects more accurately than existing approaches.

Observations of Shallow Methane Bubble Emissions From Cascadia Margin

Open questions exist about whether methane emitted from active seafloor seeps reaches the surface ocean to be subsequently ventilated to the atmosphere. Water depth variability, coupled with the transient nature of methane bubble plumes, adds complexity to examining these questions. Little data exist which trace methane transport from release at a seep into the water column. Here, we demonstrate a coupled technological approach for examining methane transport, combining multibeam sonar, a field-portable laser-based spectrometer, and the ChemYak, a robotic surface kayak, at two shallow (<75 m depth) seep sites on the Cascadia Margin. We demonstrate the presence of elevated methane (above the methane equilibration concentration with the atmosphere) throughout the water column. We observe areas of elevated dissolved methane at the surface, suggesting that at these shallow seep sites, methane is reaching the air-sea interface and is being emitted to the atmosphere.

Autonomous Vehicles Mapping Plitvice Lakes National Park, Croatia

Plitvice Lakes National Park is the largest national park in Croatia and also the oldest from 1949. It was added to the UNESCO World Natural Heritage List in 1979, due to the unique physicochemical and biological conditions that have led to the creation of 16 named and several smaller unnamed lakes, which are cascading one into the next. Previous scientific research proved that the increased amount of dissolved organic matter (pollution) stops the travertine processes on Plitvice Lakes. Therefore, this complex, dynamic but also fragile geological, biological and hydrological system required a comprehensive limnological survey. Thirteen of the sixteen lakes mentioned above were initially surveyed from the air by an unmanned aircraft equipped with a survey grade GNSS and a full frame high-resolution full-screen camera. From these recordings, a georeferenced, high-resolution orthophoto was generated, on which the following surveys by a multibeam sonar depended. It is important to mention that this was the first time that these lakes had ever been surveyed both with the multibeam sonar technique and with such a high-resolution camera. Due to the fact that these thirteen lakes are difficult to reach and often too shallow for a boat-mounted sonar, a special autonomous surface vehicle was developed. The lakes were surveyed by the autonomous surface vehicle mounted with a multibeam sonar to create detailed bathymetric models of the lakes. The missions were planned for the surface vehicle based on the orthophoto from the preliminary studies. A detailed description of the methodology used to survey the different lakes is given here. In addition, the resulting high-resolution bathymetric maps are presented and analysed together with an overview of average, maximum depths and number of data points. Numerous interesting depressions, which are phenomena consistent with previous studies of Plitvice Lakes, are noted at the lake beds and their causes are discussed. This study shows the huge potential of remote sensing technologies integrated into autonomous vehicles in terms of much faster surveys, several orders of magnitude more data points (compared to manual surveys of a few decades ago), as well as data accuracy, precision and georeferencing.

Automatic Classification of Biological Targets in a Tidal Channel Using a Multibeam Sonar

Multibeam sonars are widely used for environmental monitoring of fauna at marine renewable energy sites. However, they can rapidly accrue vast volumes of data, which poses a challenge for data processing. Here, using data from a deployment in a tidal channel with peak currents of 1–2 m s−1, we demonstrate the data-reduction benefits of real-time automatic classification of targets detected and tracked in multibeam sonar data. First, we evaluate classification capabilities for three machine learning algorithms: random forests, support vector machines, and k-nearest neighbors. For each algorithm, a hill-climbing search optimizes a set of hand-engineered attributes that describe tracked targets. The random forest algorithm is found to be most effective—in postprocessing, discriminating between biological and nonbiological targets with a recall rate of 0.97 and a precision of 0.60. In addition, 89% of biological targets are correctly classified as either seals, diving birds, fish schools, or small targets. Model dependence on the volume of training data is evaluated. Second, a real-time implementation of the model is shown to distinguish between biological targets and nonbiological targets with nearly the same performance as in postprocessing. From this, we make general recommendations for implementing real-time classification of biological targets in multibeam sonar data and the transferability of trained models.

Management and analysis of high resolution multibeam sonar surveys for geometry characterization of riverbed material

<p>The geometric characterization of riverbed material is fundamental piece of information for the management of river basins because it allows, for example, the determination of bed-load and hydrodynamics roughness and the study of geo-morphological phenomenona.<br>However information such the grading curve are not easily achievable by means of traditional field sampling methods, mostly intrusive, and to the hydraulic conditions of rivers that may have high water levels and strong flows.</p><p>Multibeam sonars represent an important alternative to traditional survey methods. Nowadays, thanks to advanced scientific knowledge, it is possible to make full use of an equipment increasingly accurate and precise. State of the art solutions have dimensions compact enough to be installed on remotly piloted vehicles and allow to obtained high resolution digital surface models of river beds. The feasibility of having models of such quality and the possibility to conduct surveys more frequently, allowing the monitoring of sedimentation and erosion phenomena as well as the dynamics of the armouring layer, have motivated the development of advanced and innovative technology to analyse these models.</p><p>The aim of this work is the development of a workflow that provides an effective method to characterize riverbed material. In order to achieve this target we start from an advanced and original survey technique, that allows to obtain high resolution digital surface models, and use an appropriate post-processing procedure.<br>We introduce first some results obtained from the analysis of digital surface models produced in laboratory or relative to well known site. In particular advanced techniques for the study of 3D model and the detection and geometric characterization of forms are investigated.<br>Then we present some data acquired at high resolution (few centimeters) with a multibeam sonar mounted on a remote controlled vessel. Field surveys were conducted in real fluvial environment with the aim of produce qualitative and quantitative information about the surface layer of riverbed.<br>Even considering some sources of uncertainty that may be present from field survey to modeling, the obtained results show how it is possible to identify and geometrically characterize several of the forms present on the surfaces analyzed. </p>