Open AccessFast ship contour extraction in SAR images
Author(s) -
Nie YuanYing,
Fan ShuChen,
Shui PengLang
Publication year - 2019
Publication title -
the journal of engineering
Language(s) - English
Resource type - Journals
ISSN - 2051-3305
DOI - 10.1049/joe.2019.0384
Subject(s) - synthetic aperture radar , artificial intelligence , computer vision , computer science , salient , minimum bounding box , active contour model , contour line , bounding overwatch , pattern recognition (psychology) , image (mathematics) , image segmentation , geography , meteorology
Fast ship contour extraction in synthetic aperture radar (SAR) images plays an important role in sea surface surveillance. Ship contours and sizes and shapes extracted from them are salient features in ship classification and recognition. This study presents a two‐step method to fast extract ship contours. At the first step, an existing object detection method is used to extract bounding boxes of ships. Next, the morphological operators are applied to each box to yield a compact external contour. At the second step, using the compact external contour as the initial contour, the active contour algorithm is used to iteratively generate ship contours. Owing to compact initial external contour, the active contour algorithm avoids the topological merging and splitting, two time‐consuming operations. In this way, ship contours can be fast extracted.