摘要
介绍了扩展分形 ( Extended fractal)技术在合成孔径雷达 ( SAR)图像特定尺寸目标检测中的应用。与传统的检测特征相比 ,扩展分形特征不仅可以依靠对比度从背景中识别目标 ,而且对目标物体的尺寸敏感。文中对扩展分形的定义以及扩展分形的均值和方差进行了分析 ,表明了扩展分形对目标尺寸和对比度的敏感特性。在扩展分形定义中 ,通过比较不同的能量增长间隔比值对扩展分形均值和方差的影响 ,确定了最优能量增长间隔比值 ,将扩展分形技术用于对实测 SAR图像中特定尺寸目标检测 ,得到了良好的识别效果。实验表明 。
The application of extended fractal (EF) features in target detection of the synthetic aperture radar (SAR) image is introduced. Compared with traditional detection features, the EF feature can identify the targets by both the contrast and the target′s size. The analyses of the definition, the mean and variance of EF prove that EF feature is sensitive to both the contrast and the size of objects. The optimum lag ratio in the definition of EF is acquired compared variable lag ratios with of influences on the mean and variance of EF. Experiment demonstrates that the EF features on the SAR image have excellent characteristics for the target sized objects.
出处
《南京航空航天大学学报》
EI
CAS
CSCD
北大核心
2004年第3期378-382,共5页
Journal of Nanjing University of Aeronautics & Astronautics
关键词
目标检测
扩展分形
合成孔径雷达
target detection
extended fractal (EF)
synthetic aperture radar (SAR)