Extracting Quantitative Information on Coastal Ice Dynamics and Ice Hazard Events From Marine Radar Digital Imagery

Abstract

Marine radars have been employed to gather data in applications that require near-continuous monitoring and tracking of objects over a wide area from a single viewpoint, independent of weather and light conditions. However, little attention has been paid toward utilizing such systems for the study of long-term phenomena and detecting anomalous environmental events or hazards that may occur infrequently but have potentially significant impacts on coastal populations. In this paper, we concentrate on tracking features in seasonally ice-covered Arctic coastal ocean environments. We have developed tools for automated analysis of ground-based radar images of landfast ice and moving sea ice to extract ice-floe trajectories and velocity fields, delineate the boundary of stable landfast ice, detect events relevant to coastal populations and identify surface vessels. We employ dense and feature-based optical flow approaches to compute motion fields from the images, active contours for delineation of stable landfast ice, and Hidden Markov Models for machine learning based event detection. We present results from the analysis of sample images jointly with a quantitative evaluation of algorithm performance relative to operator-based assessments.

A Coherent on Receive X-Band Marine Radar for Ocean Observations

“…The mean amplitude of 50 rotations of MR data are depicted in Figure 10 a and the confidence calculated from Equation (5) in Figure 10 b. At the acquisition time the ocean surface was very calm and no waves were imaged by the radar, which makes the distinction between sea ice and open water from the radar backscatter fairly easy [ 39 ]. This can also be achieved by looking into the confidence, which is a measure of the variation of the Doppler speed and therefore can be utilized to detect the speed fluctuations over open water on short time scales (rotation to rotation) in comparison to those over the sea ice.…”

Landfast sea ice breakouts: Stabilizing ice features, oceanic and atmospheric forcing at Barrow, Alaska

“…The problem of tracking ice features through radar imagery is challenging 320 due to the characteristics of the data gathered. The area within the range of the 321 radar may not be filled with ice, or may appear to not be filled with ice due to 322 occlusions that also pose a problem to traditional tracking algorithms for radars 323 (MV et al, 2013). Due to occlusions and non-rigid body deformation, traditional 324 feature tracking methods are not always effective in tracking sea ice from marine 325 radar imagery.…”

“…Due to occlusions and non-rigid body deformation, traditional 324 feature tracking methods are not always effective in tracking sea ice from marine 325 radar imagery. Instead, a combination of existing and newly developed methods is 326 used (MV et al, 2013). For areas of high ice concentration, a normalized cross 327 correlation algorithm is employed to estimate the motion field.…”

“…839 radar imagery. The average error in tracking sea ice is 1.8 pixels per eight minutes 840 (MV et al, 2013). This corresponds to a velocity of 0.079 m s -1 , although there are 841 also uncertainties in the spatial scale of the radar pixel size and timing of the radar 842 image capture.…”

Local Scale-Guided Hierarchical Region Merging and Further Over- and Under-Segmentation Processing for Hybrid Remote Sensing Image Segmentation

“…With the increasing availability of detailed geo-object information in medium-and high-resolution remote sensing images, geographic object-based image analysis (GEOBIA) [1,2] has become a new paradigm in extracting various geo-objects from these data, such as farmland division [3][4][5][6], building detection [7,8], coastal geo-object recognition [9,10] and change detection [11,12] . The advantages of the low spectral variation within geo-objects and the full utilization of textural features and shape concepts of geo-objects make GEOBIA yield more accurate image classification than the traditional pixel-based analysis method [13][14][15].…”