摘要
基于交叉熵提出了一种新的参数来衡量河流区域与其他区域的区别。该参数能在河流区域很好地进行舰船检测与桥梁检测。通过确定河流的边界,舰船和桥梁只在河流区域进行检测,所以绝大部分的虚警点都被消除。利用日本新泻地区的X波段全极化数据进行验证,表明该方法具有良好的检测性能。
A cross-entropy based parameter is proposed to measure the difference between a river area and other areas. This parameter can be employed for ship detection and bridge detection from a polarimetric SAR image. After the boundaries of the river area are determined, ships and bridges are detected only in the river area instead of the whole area. Therefore, most of the false alarms are eliminated. Finally, the performance of this new approach is validated by the X-band full polarimetric SAR data in Niigata area of Japan.
出处
《系统工程与电子技术》
EI
CSCD
北大核心
2006年第3期339-341,共3页
Systems Engineering and Electronics
基金
国家重大研究计划项目(2001CB309401)
国家自然科学基金(40271077)
清华大学基础研究基金资助课题
关键词
目标检测
极化
合成孔径雷达
特征提取
target detection
polarimetry
synthetic aperture radar
feature extraction