RTV-SIFT: Harnessing Structure Information for Robust Optical and SAR Image Registration

Registration of optical and synthetic aperture radar (SAR) images is challenging because extracting located identically unique features on both are tricky. This paper proposes a novel SAR image registration method based relative total variation (RTV) scale-invariant feature transform (SIFT), named RTV-SIFT, to extract points the edges structures construct structural edge descriptors improve accuracy. First, RTV-Harris point detection by combining RTV multiscale Harris algorithm proposed images’ significant structures. ensures high repetition rate points. Second, constructed enhanced phase congruency (EPCE), which combines Sobel operator maximum moment (PC) from structured that enhance robustness nonlinear intensity differences speckle noise. Finally, after coarse registration, position orientation Euclidean distance (POED) between utilized achieve fine matching The experimental results demonstrate superiority RTV-SIFT in different scenes capture conditions, indicating its effectiveness registration.