Detection of Small and Medium Sized Ships in Satellite Images Using YOLO Models
Yazarlar (3)
| Makale Türü |
Özgün Makale
(Diğer hakemli uluslarası dergilerde yayınlanan tam makale)
|
||
| Dergi Adı | Current Trends in Computing | ||
| Dergi ISSN | 2980-3152 | ||
| Dergi Tarandığı Indeksler | CrossRef, Road, Google Scholar | ||
| Makale Dili | Türkçe | Basım Tarihi | 06-2025 |
| Cilt / Sayı / Sayfa | 3 / 1 / 17–27 | DOI | 10.71074/CTC.1636434 |
| Makale Linki | https://doi.org/10.71074/ctc.1636434 | ||
| UAK Araştırma Alanları |
Makine Öğrenmesi
Görüntü İşleme
|
||
| Özet |
| Ship detection in satellite images is an essential part of maritime security and surveillance. This work presents a helper tool for its growth. The three models presented in this paper use YOLOv8 and YOLOv5 with colored and grayscale satellite photos to recognize small and medium-sized ships, which are frequently hard to spot in satellite photographs. Performance parameters such as mean average precision (mAP), recall, and accuracy were assessed. The models’ accuracy and recall ranged from 0.87 to 0.93 and 0.84 to 0.86, respectively, according to the results. This work uncovered a comparable performance for both grayscale and colored images and acceptable detection accuracy. In satellite images, Model 2 and Model 3 demonstrated efficacy in identifying small and medium-sized ships. |
| Anahtar Kelimeler |
BM Sürdürülebilir Kalkınma Amaçları
| Atıf Sayıları | |
| Google Scholar | 1 |