高分辨率SAR图像桥梁自动提取算法  

Automatic bridge extraction algorithm from high-resolution SAR image

作　　者：赵冠雄 沈汀[1] 

机构地区：[1]中国科学院遥感与数字地球研究所,北京100094 [2]中国科学院大学,北京100049

出　　处：《计算机工程与设计》2014年第8期2793-2797,共5页Computer Engineering and Design

摘　　要：针对合成孔径雷达(synthetic aperture radar,SAR)图像中桥梁的特征,提出一种基于高分辨率SAR图像的桥梁自动提取算法。利用改进的Otsu阈值法将图像二值化,针对面积属性,利用连通域标记法和Otsu算法思想自适应地剔除非河流区域;通过提取河流主干线与河流边界的交点初步锁定桥梁的位置与搜索区域;利用边界集合间最小距离法提取出桥梁边界集合,通过最小二乘法拟合出桥梁两侧的边界直线。对真实的SAR图像进行处理并取得了良好的提取效果。A bridge extraction method from high resolution SAR （synthetic aperture radar） images was presented. Firstly, the improved Otsu thresh method was employed to segment the image to a binary image with the river and the background. Secondly, the connected domain notation method was used to mark the connected regions and to compute their area, and then the Otsu idea was applied to the area property instead of the grayscale so as to reject the false alarm area automatically. Thirdly, the bridge location was confirmed by extracting the river trunk and boundary. Finally, the bridge edge was extracted using the raindistance of the river boundary, and two straight lines of the bridge boundary were fitted using least square estimation. Experimental results on obtained real SAR images confirm the effectiveness of the proposed method.

关 键 词：合成孔径雷达 最大类间方差 图像分割 桥梁提取 模式识别 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]