摘要
本文针对合成孔径雷达/惯性导航(SAR/INS)组合导航系统对景象匹配算法精确性、实时性和鲁棒性的要求,提出基于尺度不变特征变换(SIFT)的亚像素精度快速景象匹配改进算法。通过主成分分析后,对SIFT描述结果进行降维,提高了景象匹配的实时性。仿真试验结果表明,本算法具备实时性、优越的抗噪声、抗光照变化和抗旋转尺度变换能力。
The Synthetic Aperture Radar/Inertial Navigation System (SAPdINS) system requires real-time scene matching algorithms wich are highly precise and robust.We present a fast and subpixel-preeise registration algorithm based on extension of Scale Invariant Feature Transform (SIFT). The matching is speeded up as dimension-reduced by Principal Component Analysis (PCA). The results we obtained in tests demonstrate that the algorithm is real-time, nosie-robust and invariant in scale, orientation and illumination.
出处
《电子设计工程》
2013年第18期139-141,共3页
Electronic Design Engineering
关键词
组合导航
景象匹配
图像配准
尺度不变特征变换
integrated navigation
scene matching
image registration
Scale Invariant Feature Transform (SIFT)