改进SA方法的卫星视频图像超分辨率重建

摘要

为了提高卫星视频图像的质量,针对SA方法中无法实现小数倍超分重建和"黑色格网"问题,提出了改进的SA超分辨率重建方法.对于小数倍超分问题,首先建立参考帧低分辨率图像坐标系,以超分倍数构建参考帧高分格网,按表达的物方元素一致的原则,把所有低分辨率图像和高分辨率图像的坐标系统一在一个参考坐标系下;对于"黑色格网"改善问题,在使用运动估计参数时,为了减少由于运动补偿造成的误差,取消传统SA方法中四舍五入的运动参数补偿方法,改为以待求点为中心确定容许误差,通过容许误差选择参估点,用参估点计算高分辨率网格像素,以减少"黑色格网"现象.利用仿真和真实的卫星视频数据实验验证了该方法的有效性.%In order to improve the quality of satellite video image,an improved SA super-resolution reconstruction algorithm is proposed for the problem that the fractional-times reconstruction and the "black grid" cannot be achieved in the traditional SA algorithm.For the reconstruction problem of fractional factors,the low resolution coordinate system and high resolution coordinate system are established according to the reconstruction factors.Both coordinates are unified in one coordinate system according to the expression of the elements of the same principle.To improve the problem of "black grid",the rounding motion parameter compensation method in the traditional SA method is canceled in order to reduce the error caused by motion compensation,and the point to be determined is replaced as the center to determine the allowable error.The reference point is selected by the allowable error,then the high-resolution grid pixels are calculated by the reference point to reduce the "black grid" phenomenon.The experimental results show that the method is effective by using the simulation and real satellite video data.