Detecting inshore ships in satellite images Uydu Görüntülerinden Kiyidaki Gemi Tespiti
2015 23rd Signal Processing and Communications Applications Conference, SIU 2015, Malatya, Türkiye, 16 - 19 Mayıs 2015, ss.1220-1223, (Tam Metin Bildiri)
- Yayın Türü: Bildiri / Tam Metin Bildiri
- Doi Numarası: 10.1109/siu.2015.7130057
- Basıldığı Şehir: Malatya
- Basıldığı Ülke: Türkiye
- Sayfa Sayıları: ss.1220-1223
- Anahtar Kelimeler: Contour following, Ellipse fitting, Satellite images, Ship detection
- İstanbul Medipol Üniversitesi Adresli: Hayır
Özet
Ship detection in satellite images is used for monitoring illegal fishing and violation of coastal waters or general maritime management. SAR images has been extensively used in this manner. Recently, researchers have started using optical satellite images for ship detection. These studies can be divided into two categories as inshore and offshore ship detection. Offshore ship detection is a relatively easy problem. But, inshore ship detection is a challenging problem in which ships are located closely. In this study, we handle this problem and propose a novel method assuming we have the harbor information beforehand. In the method, the mask is morphologically thickened in single steps and adjacent ships are detected with a contour following method. Moreover, a shape model for the ships is used to eliminate false alarms. We tested the proposed method on high resolution satellite images and achieved promising results.