自适应阈值的视觉注意模型SAR舰船检测算法
发布时间：2018-11-22 08:28
【摘要】：为了解决SAR图像基于人类视觉注意模型舰船检测算法中需要人工确定经验阈值的问题,提出一种自适应阈值的视觉注意模型SAR舰船检测算法。引入最大类间方差(OTSU)法确定自适应阈值进行图像初分割,再应用视觉注意模型得到视觉显著图,最终根据显著图的统计特性进行自适应阈值分割检测出舰船目标。该算法相对于已有的视觉注意模型舰船检测算法自动化程度更高,与视觉注意模型舰船检测算法以及目前普遍使用的双参数CFAR、K-CFAR、KSW双阈值算法同时处理3种星载SAR数据——ENVISAT ASAR(25 m)、Sentinel-1(10m)和Cosmo-SkyMed(3m),进行对比分析实验,实验结果证明该算法简单、准确、高效。
[Abstract]:In order to solve the problem of manually determining the empirical threshold in the ship detection algorithm of SAR image based on human visual attention model, a visual attention model SAR ship detection algorithm based on adaptive threshold is proposed. The maximum inter-class variance (OTSU) method is introduced to determine the adaptive threshold for initial image segmentation, and then the visual salient map is obtained by using visual attention model. Finally, according to the statistical characteristics of the salience map, the adaptive threshold segmentation is carried out to detect the ship target. Compared with the existing visual attention model ship detection algorithm, this algorithm has a higher degree of automation, compared with the visual attention model ship detection algorithm and the widely used two-parameter CFAR,K-CFAR,. Three kinds of spaceborne SAR data, ENVISAT ASAR (25 m), Sentinel-1 (10m) and Cosmo-SkyMed (3m), are processed simultaneously by KSW double threshold algorithm. The experimental results show that the algorithm is simple, accurate and efficient.
【作者单位】： 山东科技大学测绘科学与工程学院;中国测绘科学研究院;
【基金】：国家基础测绘科技计划(2016KJ0103) 中国博士后科学基金资助项目(2016M591219)
【分类号】：TN957.52
本文编号：2348694
[Abstract]:In order to solve the problem of manually determining the empirical threshold in the ship detection algorithm of SAR image based on human visual attention model, a visual attention model SAR ship detection algorithm based on adaptive threshold is proposed. The maximum inter-class variance (OTSU) method is introduced to determine the adaptive threshold for initial image segmentation, and then the visual salient map is obtained by using visual attention model. Finally, according to the statistical characteristics of the salience map, the adaptive threshold segmentation is carried out to detect the ship target. Compared with the existing visual attention model ship detection algorithm, this algorithm has a higher degree of automation, compared with the visual attention model ship detection algorithm and the widely used two-parameter CFAR,K-CFAR,. Three kinds of spaceborne SAR data, ENVISAT ASAR (25 m), Sentinel-1 (10m) and Cosmo-SkyMed (3m), are processed simultaneously by KSW double threshold algorithm. The experimental results show that the algorithm is simple, accurate and efficient.
【作者单位】： 山东科技大学测绘科学与工程学院;中国测绘科学研究院;
【基金】：国家基础测绘科技计划(2016KJ0103) 中国博士后科学基金资助项目(2016M591219)
【分类号】：TN957.52
【相似文献】
相关期刊论文 前2条
1 娄智;邓浩;陈香;姚博;杨基海;;基于自适应阈值处理的表面肌电信号小波去噪研究[J];生物医学工程学杂志;2014年04期
2 田玉静;董玉民;左红伟;;时频结合自适应阈值小波包消噪算法[J];应用声学;2010年04期
相关硕士学位论文 前1条
1 钱一名;航空数据链的认知抗干扰技术研究与FPGA实现[D];北京工业大学;2016年
，本文编号：2348694
本文链接：https://www.wllwen.com/kejilunwen/xinxigongchenglunwen/2348694.html
