Matching Between SAR Images and Optical Images based on HOG descriptor
Qing Li, Zengliang Li, Guangzhou Qu · 2013
Because of the particularity of matching optical image to synthetic aperture radar image, this paper proposes an image matching algorithm for matching optical image to SAR image. This algorithm is based on Histograms of Oriented Gradient(HOG) descriptors. Firstly, we do the image preprocessing to extract prominent features. Secondly, we use the HOG descriptors as the feature space of the image matching and choose normalization cross correlation(NCC) as the measure standard of the feature matching. The experimental results show that the algorithm not only increases the probability of matching after using the effective gradient operator and feature dimensions, but also has robustness to different image regions. (4 pages)