摘要
探讨了一种优化的多源遥感影像的高精度配准算法.利用梯度算子结合Forstner算子快速提取特征点,同时采用基于熵的格网技术控制特征点的分布均匀度,在少量准确控制点的基础上以不变矩相似度量为匹配准则,采用整体松弛法匹配策略快速配准得到同名点,并利用二次多项式模型来剔除误配准点.结果表明,该方法配准速度快,得到的同名点精度高,分布均匀,可以满足遥感影像的融合与快速更新等后续处理的要求.
An optimized high-precision registering method of multi-source remote sensing images is discussed. The feature points are quickly extracted by using the operators of gradient and Forstner, and the grid technique based on the entropy is adopted to control the homogeneity of feature points. Based on some exact control points and the matching rule of the invariant moment similarity measurement and the matching strategy of the global relaxation, the homonymy points are obtained by quickly regis- tering the remote sensing images. Then the error points are eliminated by using the quadratic polynomial model. Experiment results show that the method, with quick registering speed and highly accurate and evenly distributing points, can meet the need of the image fusion and quick updating of the remote sensing images.
出处
《同济大学学报(自然科学版)》
EI
CAS
CSCD
北大核心
2008年第10期1427-1432,共6页
Journal of Tongji University:Natural Science
关键词
遥感影像
配准
特征点
不变矩
松弛匹配
remote sensing images
registration
feature point
invariant moment
relaxation matching