SAR real-time guidance system based on multi-scale FAST-BRISK

Abstract

In recent years, SAR (Synthetic Aperture Radar) has been widely used in the real-time guidance system. As the foundation of system, SAR image registration directly affects the guidance performance. In order to achieve the high precision and low computation, in this paper, we adopt well-known FAST (Features from Accelerated Segment Test) for feature detection and BRISK (Binary Robust Invariant Scalable Keypoints) for feature description. Then, we use Hamming distance and RANSAC (Random Sample Consensus) to match key points and estimate transformation parameters. To evaluate the performance of algorithms, three simulation experiments have been accomplished. The results and comparative analysis show a better performance of our proposed method.