中文摘要：

提出了一种基于联合像素模型和信号子空间加权平均方法（weighted average of signal subspaces,WAVES）的多频带In-SAR自适应相位融合方法。首先,采用联合像素模型对各频带InSAR信号建模,然后,借鉴WAVES的思想对多频带信号子空间进行聚焦和加权,获得通用协方差矩阵,最后,利用窄带估计方法估计出绝对干涉相位。该方法可以在图像配准精度很差（可以允许达到一个分辨单元）的条件下获得稳健的干涉相位估计性能,仿真数据和SIR-C实际雷达数据验证了此方法的有效性。

英文摘要：

An interferometric phase fusion method for multi-frequency InSAR systems is proposed in this paper.The method takes advantage of the coherence information of neighboring pixel pairs to automatically coregister the SAR images in difference frequency bins.Using WAVES（weighted average of signal subspaces）,the method transforms the signal subspaces at many frequency bins into a general correlation matrix at one focusing frequency using a transformation matrix and a weighting matrix;then estimates the elevation using narrow-band method.The method can provide more accurate estimation of the interferometric phase even when the coregistration error reaches one pixel than those obtained with previous methods.The effectiveness and robustness of the method is verified by simulated data and SIR-C real data.