Image Registration Algorithm in SAR Image with Linear Features
Jing Zeng · International Journal of Applied Mathematics & Statistics/International journal of applied mathematics and statistics · 2013
Typical artificial target with visible linear features and synthetic aperture radar (SAR) image can be the objects of study. It is proposed that an adaptive multi-scale fast Beamlet transformation method can extract artificial target in the visible and common features of SAR image-line features. They are based on the linear characteristic structure control point. They are designing a matching function based on the characteristics of the control point. They are based on rough registration. It is consistent and accurate registration method based on control points. We can treat automatic image registration from coarse to fine. The experimental results can show the better result in SAR images with visible large gradation differences. The rotations and translations are still able to precise alignment.