期刊名称：International Journal of Advanced Information Technology
印刷版ISSN：2231-1920
电子版ISSN：2231-1548
出版年度：2011
卷号：1
期号：5
出版社：Academy & Industry Research Collaboration Center (AIRCC)
摘要：Maritime security includes reliable identification of ship entering and leaving a nation's territorial waters. Sea target detection from remote sensing imagery is very important, with a wide array of applications in areas such as fishery management, vessel traffic services, and naval warfare. Automated systems that could identify a ship could complement existing electronic signal identification methods. A new classification approach using shape and texture is introduced for Ship detection. Texture information can improve classification performance. This approach uses both shape and texture features. Feature extraction is done by Hu's moment invariants with several classification algorithms. This paper presents an overview of automatic ship recognition methods with a view towards embedded implementation on optical smart cameras. Therefore this approach may attain a good classification rate
关键词：shape and texture; Hu's moment invariants; ship recognition; smart cameras
