| 摘要: |
| 针对合成孔径雷达（Synthetic Aperture Radar，SAR）图像匹配中虚假匹配和鲁棒性欠缺的问题，提出了一种鲁棒的SAR图像匹配方法。首先通过正射校正来进行预处理，从而消除SAR图像的旋转；然后利用相关匹配方法初步提取同名点；最后根据同名点的空间位置来筛选同名点并计算仿射矩阵进行匹配，提高了SAR图像匹配的精度和鲁棒性。实验证明了所提方法的有效性。 |
| 关键词: SAR图像匹配 相关匹配 尺度不变特征变换 |
| DOI： |
|
| 基金项目: |
|
| A robust SAR image matching algorithm |
| WU Yuan,QIAO Wensheng,SHAO Peng |
| (Southwest China Institute of Electronic Technology,Chengdu 610036,China) |
| Abstract: |
| To solve the fake matching and less reliability problems among synthetic aperture radar(SAR) image matching algorithms,a robust SAR image matching algorithm is proposed.It removes the rotation of SAR image through ortho-rectification,then computes the homonymy points through correlation method,and filters the homonymy points by their positions to compute the affine matrix,thus improving the accuracy and reliability.The experiment results show that the proposed algorithm is effective. |
| Key words: SAR image matching correlation matching scale-invariant feature transform(SIFT) |
