GGSOR: A Gaussian-Gamma-Shaped bi-windows based descriptor for optical and SAR images matching

A matching method for optical and synthetic aperture radar (SAR) images, robust to speckle noise, is presented. Firstly, a coarse correction to eliminate rotation and scale change between images is performed. Secondly, features robust to speckle noise of SAR image are detected by improving the original phase congruency based method. Then, feature descriptors are constructed by combining the Gaussian-Gamma-Shaped bi-windows based gradient operator and the histogram of oriented gradient pattern. Finally, descriptor similarity and geometrical relationship are combined to constrain the matching processing. The experimental results demonstrate that the proposed method provides significant improvement in correct matches number and image registration accuracy compared with other traditional methods.