학술논문

Ship wake detection in SAR images: a segmentation-based approach
Document Type
Conference
Source
IEEE International Geoscience and Remote Sensing Symposium Geoscience and remote sensing symposium Geoscience and Remote Sensing Symposium, 2002. IGARSS '02. 2002 IEEE International. 1:110-112 vol.1 2002
Subject
Geoscience
Signal Processing and Analysis
Marine vehicles
Image segmentation
Detectors
Traffic control
Data mining
Robust control
Computer vision
Motion detection
Image sensors
Sensor phenomena and characterization
Language
Abstract
This paper deals with the derivation of a fully automatic and reliable ship wake detector (SWD), which can operate in conjunction with a ship detector scheme (SD) for ship traffic control against non-homogeneous sea background in SAR images. A novel segmentation-based technique is introduced for ship wake detection, which overcomes the drawbacks suffered by standard techniques operating in non-homogeneous background. This is composed of a preliminary segmentation stage to extract context information, followed by a modified wake detector. The new SWD scheme is shown to yield a large performance improvement on both simulated and real SAR data.