A Sea-Ice Lead Detection Algorithm for Use With High-Resolution Airborne Visible Imagery

摘要: The detection of leads, or cracks, in sea ice is critical for the derivation sea-ice freeboard from altimetric measurements elevation. We present an approach lead using high-resolution visible imagery airborne platforms. develop a new algorithm, i.e., algorithm minimal signal (SILDAMS), that detects clouds, extracts and classifies types within leads imagery. Cloud based on assessment local variances pixel brightness across image scenes where available coincident are used to confirm suspected cloudy scenes. extraction step computes affine time-frequency distributions (minimal signal) Red, Green, Blue channels each image. transformed outputs combined take advantage three simultaneously. Finally, geolocations extracted set uniform thresholds typing (including open water, thin ice, gray ice) along flight line. SILDAMS was tested data Digital Mapping System (DMS). DMS digital photographs represent highest resolution ( ≈10 cm) over were collected during NASA Operation IceBridge flights Antarctic Arctic 2009 2010, respectively. demonstrate has high capability 99%.