摘要
图像配准是INSAR技术中关键的一环。可见光图像配准中的相位相关法,因为对噪声具有很高的稳健性,使用相当普遍。针对SAR图像特点,在相位相关法原有基础上,提出几点改进。特别是边缘模糊方面,从信号处理原理角度进行了深入的分析,并且在MATLAB环境下仿真得到验证。通过在原有相位相关法的基础上的几处改进,配准的速度和精度有很大提高,满足了INSAR对两幅SAR图象配准的精度要求,为INSAR下步的相位解缠提供了很好的基础。实验证明了该方法的有效性,可以作为INSAR中关键的一步。
Image registration is an important part of INSAR. Phase correlation is used widely in image registration of visible light images because of its steadiness. In this paper, according to the characters of SAR images, several improvements are proposed. Especially in the part of edge bluring, deep analyzing is proposed from the signal processing prospect. And it is identified in the environment of MATLAB. Through several improvements based on former pahse correalation, the speed and accuracy are improved obviorsly, which can meet the requirement of INSAR registration and are good to the next work of INSAR. The experiment shows the effectiveness of the method.
出处
《计算机仿真》
CSCD
2007年第9期210-213,共4页
Computer Simulation
