文章摘要
颜普,唐庆伟,苏亮亮.基于面积比不变量的遥感图像配准算法[J].安徽建筑大学学报,2022,30():
基于面积比不变量的遥感图像配准算法
Remote Sensing Image Registration Using Area Ratio Invariance
投稿时间:2021-05-24  修订日期:2021-07-25
DOI:
中文关键词: 图像配准  遥感图像  面积比不变量  中值K最近邻图
英文关键词: image registration  remote sensing image  area ratio invariance  median K nearest neighbour graph
基金项目:安徽高校协同创新资助项目(GXXT-2019-007);国家自然科学基金青年基金资助项目(61901006);安徽省自然科学基金青年项目资助项目(1908085QF281);安徽省重点研究与开发计划资助项目(202004a07020050);安徽建筑大学博士启动基金(2017QD13)
作者单位E-mail
颜普* 安徽建筑大学 yanpu8188@163.com 
唐庆伟 安徽建筑大学  
苏亮亮 安徽建筑大学  
摘要点击次数: 1779
全文下载次数: 718
中文摘要:
      摘要:为提高遥感图像配准性能和效率,提出一种基于面积比不变量的遥感图像配准算法,其目的是利用面积比这一邻域约束来快速去除错误匹配。为表征匹配的邻域结构,对每个初始匹配构建中值K最近邻(KNN)图,在中值KNN图中计算面积比不变量作为邻域结构的表示。由于错误匹配会打破邻域结构的相似性,其邻域结构之间的误差较大,可设置阈值来将其移除,该算法最大的优点是不需要进行迭代便可以得到较为精确的特征匹配.实验结果表明,该算法具有较好的匹配效果和更快的处理速度,能够较好地实现遥感图像的自动配准。
英文摘要:
      In order to improve the performance and efficiency of remote sensing registration, a remote sensing image registration algorithm using area ratio invariance was proposed, and its goal was to quickly remove mismatches by preserving local neighborhood structures using area ratio. To represent the local neighborhood structure of a correspondence, a median K nearest neighbour (KNN) graph was constructed for each correspondence and the area ratio invariance was calculated in the median KNN graph as a representation of neighborhood structure. Since the mismatch could disturb the neighborhood structure, the error between two neighborhood structure was relatively large, and then a threshold value was set to remove the mismatch. The greatest advantage of this algorithm could accomplish the mismatch removal without iteration. Experimental results show that the proposed method has better matching performance and more faster than state-of-the-art methods, which can well realize the automatic registration of remote sensing images.
查看全文   查看/发表评论  下载PDF阅读器
关闭

分享按钮