期刊名称：Journal of Advances in Information Technology
印刷版ISSN：1798-2340
出版年度：2019
卷号：10
期号：2
页码：48-53
DOI：10.12720/jait.10.2.48-53
出版社：Academy Publisher
摘要：Space-based detection of ship targets is a significant way for fishery management, naval warfare, vessel traffic services today. Now the algorithm of ship target detection based on SAR and visible light is very mature, while the research of infrared ship target detection develops slowly as the resolution of infrared payload on orbit is relatively low. With the development of infrared technology, ship recognition based on infrared band becomes possible. As the space-borne infrared images have the characteristic of low SNR, low contrast and is easily influenced by the weather, light, sea, cloud background, a satellite-borne infrared target recognition realization is expected to develop to provide the guidance for the engineering applications. In this paper an innovative method of sea land segmentation by the statistics of power spectrum based on block combined with region growing method, background suppression by the revised TOP_HAT filter operator united with saliency map method is adopted. The measured long-wave infrared images are employed to verify the relevant method and the results show that the method proposed can detect the ship targets clearly.
