摘要
InSAR复图像配准是合成孔径雷达3维成像技术的一个关键步骤,要求达到亚像素级的配准精度。在分析了传统相关系数法的基础上,提出了改进的基于频域相关的快速配准算法。该方法分别采用频域相关和线性调频Z变换进行粗配准和精配准;最后利用傅里叶变换的平移性质对辅图像插值重采样。应用该算法对实际数据进行了实验,并从配准时间和精度两方面进行了分析。实验结果证明该方法在保证配准精度的同时,能够有效减少运算量,大大提高运算效率。
The registration for complex images is an important step on processing technology of 3D imaging in interferometric synthetic aperture radar,and it is desired to have the sub-pixel precision of registration.On the basis of analysing the traditional correlation coefficient method,a fast registration method used frequency-domain correlation coefficient and chirp Z transform respectively for coarse and fine registration was proposed,and resampled the interpolated slave image by the translation property of FT finally.Real data were chosen to perform the above mentioned algorithms,and meanwhile the runtime and precision were analyzed.The results proved that the algorithm was valid,robust and efficient.
出处
《测绘科学技术学报》
北大核心
2011年第3期204-207,共4页
Journal of Geomatics Science and Technology
基金
国家自然科学基金资助项目(60872100)
