摘要
针对合成孔径雷达(SAR)目标检测精确性、实时性和鲁棒性的要求,设计了一种基于局部窗口的SAR图像目标检测算法.该算法在对获取的SAR图像进行去噪和分割处理的基础上,基于尺度不变特征变换(SIFT)实现了亚像素精度快速配准策略;同时,通过SIFT特征的描述结果降维和基于局部窗口的最大期望算法(EM)实现了目标检测.实验结果表明,该算法对复杂背景和光照、旋转变化有较强的自适应性,获得了理想的目标检测效果.
In terms of the requirements of the target detection accuracy, real-time and robustness for synthetic aperture radar(SAR), this paper designs an algorithm for SAR image target detection based on local window. On the basis of conducting the denoising and segmentation processing to the acquired SAR image, this algorithm implements the strategy of fast registration for sub-pixel accuracy, based on SIFT. Meanwhile, it also achieves the target detection by the feature description results of SIFT lowering the dimension, and based on the expectation maximization algorithm of local window. Experimental results show that this algorithm has a better adaptation to the complex background and illumination along with rotation change, achieving an ideal effect of target detection.
出处
《空军预警学院学报》
2015年第3期169-172,176,共5页
Journal of Air Force Early Warning Academy
关键词
合成孔径雷达
SAR图像配准
尺度不变特征变换
目标检测
局部窗口
最大期望算法
synthetic aperture radar(SAR)
SAR image registration
scale invariant feature transform(SIFT)
target detection
local window
expectation maximization algorithm