广义相关导向矢量的InSAS干涉相位估计

针对干涉合成孔径声呐(InSAS)的干涉相位估计问题,利用InSAS复图像对方位向无偏移的特点,提出一种基于广义相关导向矢量模型的干涉相位估计方法.通过构造广义相关导向矢量模型,充分利用当前像素及其周围像素对的相干信息,进而采用所构建的代价函数估计出干涉相位.该方法具有图像自适应精确配准功能,能够在配准误差不超过一个分辨单元的情况下获得准确的干涉相位估计结果.仿真数据及InSAS实测数据的处理结果证明了方法的有效性.

InSAS interferometric phase estimation based on a generalized cross-correlation steering vector

According to the interferential synthetic aperture sonar(InSAS) interferometric phase estimation problem,the synthetic aperture image pair has the characteristic of zero offset in the azimuth direction.On the basis of this characteristic,a method based on the generalized cross-correlation steering vector was proposed for InSAS interferometric phase estimation.In this method,the concrete generalized cross-correlation steering vector was employed in making use of the coherence information of the neighboring image pixels,and a novel cost function was designed for phase extracting.The method can auto-coregister the SAS images and provide accurate interferometric phase estimation even if the coregistration error reaches one pixel.Results of computer simulation and the real data of InSAS verify the validity of the method.