摘要: |
当SAR图像噪声过强时,采用归一化积相关算法得到的相关面对比性较差,匹配结果
可靠性降低。为解决此问题,给出一种新的SAR图像分块匹配算法。该算法将实时图分块并
分别与基准图进行匹配,利用分块图像之间的相关性来找正确匹配点,增强了匹配结果的可
靠性。实验证实该算法在实时图有强噪声的情况下依然可以正确匹配。 |
关键词: SAR 分块匹配 归一化积相关 强噪声 |
DOI:doi:10.3969/j.issn.1001-893x.2013.06.010 |
|
基金项目:国家自然科学基金资助项目(61102166) |
|
A new block matching algorithm for SAR image |
ZHANG Dong-xing,ZHU Ming-bo,LI Xiang-ping,ZHAO La,ZHU Yong-xin |
() |
Abstract: |
When Synthetic Aperture Radar(SAR) image noise is too strong, the reliability o
f matching based on normali
zed product correlation algorithms reduces obviously. A SAR image block matching
algorithm is given to solve this problem. The relativity between the sub-block
images matching is utilized to enhance the reliability of the matching result.
The experiments confirm that the algorithm can still get correct match under the
circumstance of real-time image with strong noise. |
Key words: SAR block matching normalized product correlation strong image noise |