{"title":"Infrared and Low Light Image Registration from Coarse-to-Fine Matching","authors":"Jiahui Wang, Zhengyou Wang, W. Lu, Shanna Zhuang","doi":"10.1109/ICCEAI52939.2021.00024","DOIUrl":null,"url":null,"abstract":"At present, due to the different imaging characteristics of infrared and low light bands, they are complementary, and are widely used for multi-modal image registration and fusion. Image registration is a precondition for image fusion. For infrared and low light image registration, this paper first performs rough matching of image features based on the grid motion statistics method. Then, precision matching algorithm based on the combination of distance constraint and slope consistency is proposed, and the coarse matching feature points are initially screened for precision matching. Finally, the coarse matching after screening is selected by the random sampling consensus algorithm for the secondary screening of fine matching, and the final feature matching is obtained. The image registration strategy in this paper performs well in the evaluation indexes of accuracy and recall, which improve the accuracy of image registration.","PeriodicalId":331409,"journal":{"name":"2021 International Conference on Computer Engineering and Artificial Intelligence (ICCEAI)","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2021-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2021 International Conference on Computer Engineering and Artificial Intelligence (ICCEAI)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICCEAI52939.2021.00024","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
At present, due to the different imaging characteristics of infrared and low light bands, they are complementary, and are widely used for multi-modal image registration and fusion. Image registration is a precondition for image fusion. For infrared and low light image registration, this paper first performs rough matching of image features based on the grid motion statistics method. Then, precision matching algorithm based on the combination of distance constraint and slope consistency is proposed, and the coarse matching feature points are initially screened for precision matching. Finally, the coarse matching after screening is selected by the random sampling consensus algorithm for the secondary screening of fine matching, and the final feature matching is obtained. The image registration strategy in this paper performs well in the evaluation indexes of accuracy and recall, which improve the accuracy of image registration.