English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Sea ice is a very important issue in winter navigation in the Arctic waters and also in the Baltic Sea. The winter traffic in ice is assisted by ice breakers in ice covered areas with heavy ship traffic. In other ice covered areas, however, the vessels must have the capability to navigate in sea ice without aid, meaning that they have to be designed especially for ice conditions. Earlier e.g. the Finnish ice breakers used helicopters to find the best routes in sea ice. However, today the expensive helicopter use has been replaced by utilization of remote sensing data. And for the ships navigating in ice on their own, it is even more important to have valid and useful information of the current sea ice around the ship to find its route through ice. Especially Synthetic Aperture Radar (SAR) data are useful during the dark and often cloudy northern winters. Because C-band SAR, like Radarsat (1 and 2) and Envisat ASAR, is an active instrument it also works during the dark periods and has a suitable EM wavelength to propagate through the cloud cover. Our aim is to produce automatically generated sea ice products from SAR data, such that they are informative and easy to interpret. Sea ice types in SAR images are best described by the edges present in the images. The type and amount of edges give us information on the ice types in addition to commonly used areal backscattering values and statistics. Here we present some novel features for sea ice SAR data classification, mainly derived from detected edges. The preprocessing of the SAR data is also an important step. First the data is rectified to a map projection, Pearson (1990). We use the Mercator projection, also used in nautical charts, in the Baltic Sea and the Polar Stereographic projection for the Arctic. Also an incidence angle grid over the SAR area in appropriate projection is generated from the data included in the SAR data. After applying a land masking and incidence angle correction, Karvonen et al. (2002), using the generated incidence angle grid, we perform an edge-preserving speckle filtering. Then we apply an intensity-based segmentation, and produce a multi-resolution presentation in three resolutions by combining adjacent segments with different criteria for different resolutions. We also detect the edges and corners from the full-resolution SAR data. After the segmentation, the detected edges can be divided into two groups: segment edges (segment boundaries) and edges within segments. The segment edge features describe the contrast and shape of the segment, and the within-segment edge features give additional geophysically relevant information on the ice properties within the segment. The segment shape gives information on the segment, e.g. ice floes typically can have polygonal or round shape, and ice ridges and cracks or leads typically are narrow but long segments, 8

C-Band Sea Ice SAR Classification Based on Segmentwise Edge Features