基于特征匹配的空间目标ISAR图像横向定标

摘要

The inverse synthetic aperture radar (ISAR) image cross-range scaling is the key technique for obtaining the dimension and shape of targets, and the key point is the estimation of the rotation angle estimation. A new ISAR image cross-range scaling method based on feature registrationis presented. Firstly, adequate interested points are extracted from two adjacent ISAR images of satellite target by Harris and SIFT. Secondly, those points are pinpointed by matching with Radon Sample Consensus (RANSAC) principles. At last, the cost function of the coordinate-difference is established to estimate the rotation angle. The processing results of simulated data of satellite show that the algorithm has a good performance.%逆合成孔径雷达(Inverse Synthetic Aperture Radar,ISAR)图像横向定标是获取目标外形和尺寸信息的关键,重点是目标转角的估计.提出了一种基于特征匹配的ISAR图像横向定标方法,将回波数据等长且相邻的ISAR卫星图像用Harris和SIFT算法提取特征点,然后利用随机采样一致性进行特征点配准,最后建立坐标位置差代价函数估计目标转角,完成ISAR图像横向定标.采用卫星仿真数据对算法进行了验证,实验结果表明该算法能够准确估计横向定标因子,定标精度高.