A Bayesian Method for ViSAR Image Fusion Using Effective Reflection Coefficient

Since the isotropic scattering assumption does not hold in a wide-angle synthetic aperture radar, the video synthetic aperture radar (ViSAR), a new sensing mode by sequentially forming SAR images on a series of contiguous or overlapping sub-apertures, has the promising capability to capture aspect-dependent scattering behavior of objects. In this paper, a new Bayesian method for fusing ViSAR images is proposed. First, the notion of effective reflection coefficient is defined to characterize aspect-dependent scattering behavior of objects. Based on the ViSAR images sequence, the spatial effective reflection coefficient in the area of interest (AOI) when observing at different aspect angles are estimated. A Bayesian hypothesis testing is then derived to fuse the obtained effective reflection coefficient estimates for detecting scatterers in the AOI. The performance of the proposed method is evaluated and compared with that of the conventional GLRT-based ViSAR image fusion method in a simulated scenario. Numerical results demonstrate the superiority of the proposed method in capturing the aspect-dependent scattering characteristics as well as the spatial structure of objects.