多尺度自卷积方差显著性SAR图像目标检测  

SAR Image Target Detection Based on Multi-scale Auto-convolution Variance Saliency

作　　者：王国力[1] 周伟[1] 丛瑜[1] 关键[1] 

机构地区：[1]海军航空工程学院电子信息工程系,山东烟台264001

出　　处：《应用科学学报》2013年第6期607-612,共6页Journal of Applied Sciences

基　　金：国家自然科学基金(No.61179017;No.61201445);"泰山学者"建设工程专项经费资助

摘　　要：针对SAR图像中显著性目标检测问题,提出一种基于多尺度自卷积方差显著性的自适应检测算法.该算法在对SAR图像多尺度自卷积运算基础上,通过计算MSAV得到方差显著图.设计了一种自适应阈值检测器,完成SAR图像中显著性目标的检测.实验结果表明,在复杂背景环境下,所提算法能有效检测出与人类视觉较为一致的显著性目标.To detect salient objects in SAR image, an adaptive detection method is proposed based on multi-scale auto-convolution variance （MSAV） saliency. With multi-scale auto-convolution operation in SAR image and by calculating MSAV, a variance saliency map is obtained. An auto-threshold-selecting detector is constructed and salient object detection from the SAR image is achieved. Experimental results show that, by applying the proposed algorithm to a complex scene, salient objects consistent with human visual sense can be effectively detected.

关 键 词：合成孔径雷达图像 目标检测 多尺度自卷积 方差显著性 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置]