摘要
基于改进SURF特征匹配算法在提取出两幅图像的特征向量后,经过相似性度量进行匹配。采用RANSAC算法设定阈值,将数据中不好的"外点"剔除,提高了特征点的匹配效率。
Based on improved SURF feature matching algorithm, two feature-extracted images are matched with similarity. Random Sample Consensus （RANSAC） algorithm is used to set the threshold for removing the “outside point” from the data set. Experimental results show that the method improves the matching efficiency.
作者
张强
韩松奇
于微波
ZHANG Qiang;HAN Songqi;YU Weibo(School of Electrical ＆ Electronic Engineering,Changchun University of Technology,Changchun 130012,China)
出处
《长春工业大学学报》
CAS
2018年第4期373-377,共5页
Journal of Changchun University of Technology
基金
吉林省科技厅基金资助项目(20140204018GX)
