Ship detection in harbor area in SAR images based on constructing an accurate sea-clutter model
Autor: | Peng Shujuan, Li Zhi, Qu Changwen, Li Jianwei, Zhou Qiang, Liu Chen |
---|---|
Rok vydání: | 2017 |
Předmět: |
Pixel
Computer science business.industry ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION 0211 other engineering and technologies 02 engineering and technology Image segmentation STRIPS Constant false alarm rate law.invention law 0202 electrical engineering electronic engineering information engineering Detection performance Clutter 020201 artificial intelligence & image processing Segmentation Computer vision Artificial intelligence Fisher model business 021101 geological & geomatics engineering |
Zdroj: | 2017 2nd International Conference on Image, Vision and Computing (ICIVC). |
DOI: | 10.1109/icivc.2017.7984450 |
Popis: | Aiming at the problem of low ship detection performance of SAR image in port area, this paper proposes SAR image ship detection based on sea clutter accurate modeling in port area. In order to establish the sea clutter model accurately, this paper tries to eliminate all the pixels of land and the target of the suspected ships before modeling. Firstly, the SAR image is preprocessed by fine Lee rate ratio, and then the land mask is realized by the land-sea segmentation algorithm based on SLIC super-pixel segmentation. Secondly, the Fisher model parameters are obtained and the global threshold is obtained by dichotomy to eliminate the suspected targets. Based on the land mask and the removal and suspected ship target, the algorithm of model similarity fitting is used to self-adapt the sea clutter model distribution of SAR image in port area. Finally, to prevent from leakage, expand some size of strips on the land side of the land-sea divider line, and then conduct ship detection in with constant false alarm rate on the fitting sea clutter distribution. The simulation results verify the validity of the SAR image in ship detection in port area. |
Databáze: | OpenAIRE |
Externí odkaz: |