摘要
图像配准就是将不同时间、不同传感器或不同条件下对同一景物获取的两幅或多幅图像,进行比较找到该组图像中的共有景物,或是根据已知模式到另一幅图中寻找相应的模式。利用遥感图像进行目标监测需要进行图像配准处理。匹配算法如何达到高精度、高匹配正确率和实时性成为人们追求的目标。文章在传统匹配算法的基础上,提出两点改进:一是通过PCA-圆形SIFT算法提取图像特征角点,降低维数,优化计算;二是利用图像角点作为单调递增阈值序列SSDA算法匹配的基本像素点,利用遥感图像信息特征降低匹配计算量。最后进行实验仿真,结果表明,改进后的算法使得配准速度进一步提高,能够满足遥感图像配准实时性的要求。
Image registration is the process of matching two or more images derived from the same scene,at different time,by different sensors or at different views of angle.Target monitoring with remote sensing images requires pre-registration.How to effectively enhance the accuracy of image matching has become the key point.This paper proposes two improvements based on the traditional matching algorithms：firstly getting lower dimension and optimization calculation through PCA-SIFT algorithm;secondly reducing the amount of matching calculation according to remote sensing image features.The simulation results show that the proposed algorithm can reduce noise,greatly improving the speed of registration at the same time.
出处
《华东交通大学学报》
2013年第1期15-21,共7页
Journal of East China Jiaotong University
基金
国家自然科学基金项目(61261041)
关键词
图像配准
遥感图像
SSDA
PCA-SIFT
image registration
remote sensing image
SSDA（sequential similarity detection algorithms）
PCA-SIFT（principal component analysis-scale-invariant feature transform）
