【摘要】 针对光学基准图与SAR图像的配准问题,给出了基于边缘图粗匹配和角点精匹配的高精度匹配方法;在此基础上,基于CRLB理论推导得到光学基准图用于SAR图像匹配的性能边界值。数据仿真验证了匹配方法的有效性,分析了CRLB受噪声的影响,表明匹配方法在噪声较小时基本接近CRLB。更多还原

【Abstract】 This paper addresses registration method and registration precision problems for optical base image and SAR image.A high precision registration method composed of edge-based rough registration and corners-based fine registration is suggested.Based on Cramér-Rao Lower Bound(CRLB) theory,the fundamental performance bounds on image registration are derived.The registration experimental result proves the validity of this method.The numerical example is presented to illustrate how CRLB is influenced by noise,which also shows that the registration precision of proposed method is close to CRLB when the noise level is lower.更多还原