高分辨率合成孔径雷达图像高速公路检测法

针对高分辨率合成孔径雷达（SAR）图像中高速公路的特征，提出了一种结合多级非线性加权平均中值滤波和Hough变换的高速公路检测算法。该算法首先对原始高分辨率SAR图像进行多级非线性加权平均中值滤波，抑制斑点噪声，同时较好地保留图像的几何特性。然后对滤波后的图像进行Hough变换快速检测高速公路，并将检测到的高速公路信息叠加到原始SAR图像上显示。实验结果证明该算法能快速、有效地从不同工作模式下取得的高分辨率SAR图像中检测到直线高速公路。

Freeways detection from high-resolution SAR images

Concerning the characteristics of freeways in high-resolution Synthetic Aperture Radar (SAR) images, this paper proposed a detection algorithm for freeways. The algorithm mainly used multilevel nonlinear weighted mean median filtering algorithm and Hough transform. First, multilevel nonlinear weighted mean median filtering was run to the original high resolution SAR images. It can not only inhibit speckle noise more effectively but also preserve the geometrical features of the original SAR images. Then, Hough transform was run to filtered images to detect freeways, and superposed the detected information on the original SAR images to display. The experimental results indicate that the algorithm can detect freeways rapidly and effectively from high resolution SAR images got from different modes.