摘要: |
基于多特征点的遥感图像配准方法应用在油田遥感图像上时,很难为其找到足够多的正确对应关系且花费时间也较长。为此,提出了基于各向异性尺度空间的遥感图像配准方法。该算法首先使用自适应的侧窗滤波技术构建各向异性尺度空间,然后采用改进的邻域分块思想进行特征描述,最后使用增强的匹配点过滤增加特征点匹配的数量。实验结果表明,所提方法不仅保证了配准精度在90%以上,而且在速度上比原算法提高了10%以上。 |
关键词: 遥感图像 图像配准 各向异性 侧窗滤波 稀疏向量场共识 |
DOI: |
|
基金项目:国家自然科学基金资助项目(11701159);国家重大科技专项(2017ZX05019-005);黑龙江省自然科学基金项目(LH2019F004) |
|
Remote sensing image registration by using anisotropic scale space |
LI Hong,WANG Peng,BI Bo,TANG Jinping |
(1a.School of Electrical Information Engineering,Northeast Petroleum University,Daqing 163318,China;School of Mathematics and Statistics,Northeast Petroleum University,Daqing 163318,China;
2.School of Public Health,Hainan Medical College,Haikou 571101,China;School of Data Science and Technology,Heilongjiang University,Harbin 150080,China) |
Abstract: |
When the remote sensing image registration method based on multi-feature is applied to remote sensing images of oil fields,the number of feature point matches obtained is small,and time consumption is long.To solve the problem,a remote sensing image registration method based on anisotropic scale space is proposed.it firstly uses adaptive side window filtering technology to construct an anisotropic scale space,which significantly preserves the edges of the image.Then it uses a new type of descriptor to reduce the dimension of the feature descriptor.Finally,it uses a more robust and enhanced matching point filtering algorithm to obtain more accurate matching points and ensure the registration accuracy.The results show that this algorithm perform better than the original algorithm and other mainstream algorithms in terms of the number of correct matching pairs,matching accuracy and computational time.In particular,this algorithm guarantees that the registration accuracy is about 90% which is higher than that of the original algorithm,and the speed is increased by more than 10%. |
Key words: remote sensing image image registration anisotropic diffusion side window filter sparse vector field consensus |