Title :
Automatic detection of ship tracks in satellite imagery
Author :
Weiss, John M. ; Luo, Ruixuan ; Welch, Ronald M.
Author_Institution :
Dept. of Math. & Comput. Sci., South Dakota Sch. of Mines & Technol., Rapid City, SD, USA
Abstract :
Certain unusual cloud features visible over water in satellite images are caused by ship smoke stack pollution. Ship tracks form long, thin, complex features in satellite images. These features do not typically follow straight lines or other low-order polynomial curves, making automated detection difficult. Nonetheless, the ability to automatically detect ship tracks is an important one, with military, navigation, environmental, and rescue applications. A multi-step automated approach for detection of ship tracks in AVHRR images has been developed. First, an enhanced ship track image is produced from AVHRR channels 1, 3, and 4. Ship tracks stand out as bright linear features, or ridges, in this enhanced image. Then a new technique called ridge detection by iterated erosion is applied to this enhanced image. Finally, postprocessing based on connected components analysis is used to eliminate ridges that do not correspond to ship tracks
Keywords :
air pollution; edge detection; image enhancement; iterative methods; optical tracking; remote sensing; ships; AVHRR images; automatic detection; bright linear features; cloud feature; enhanced ship track image; multi-step automated approach; postprocessing; ridge detection by iterated erosion; ridges; satellite imagery; ship smoke stack pollution; ship tracks; Brightness; Clouds; Computer science; Image segmentation; Marine vehicles; Mathematics; Military satellites; Pollution; Satellite broadcasting; Smoke detectors;
Conference_Titel :
Geoscience and Remote Sensing, 1997. IGARSS '97. Remote Sensing - A Scientific Vision for Sustainable Development., 1997 IEEE International
Print_ISBN :
0-7803-3836-7
DOI :
10.1109/IGARSS.1997.615827