摘要
室外建筑物图像由于建筑本身纹理单一且重复、图像间的基线较宽等特点,应用传统的匹配算法在进行图像匹配时会出现正确的匹配特征点少,匹配扩散后得到的准稠密和稠密匹配特征点的匹配多义性问题,同时,误匹配也影响了建筑物的三维重建效果。针对上述问题,研究并改进了适用于建筑物图像的匹配扩散算法。应用自适应支持加权匹配度量准则减少误匹配,同时扩散过程中应用仿射传递来获取新匹配的最佳邻域窗口。实验表明,改进的算法对于建筑物图像能获得数量较多、较正确的准稠密匹配扩散结果。
Textures of outdoor buildings are locally insufficient and repetitive as well as their images are wide baseline.Previous traditional matching methods do not work well on these images,because they can t obtain enough correct matches and there are match ambiguity problems in dense matching algorithm and quasi-dense matching algorithm after matching propagation,which seriously effects 3D reconstruction model of buildings.Based on above problems,this paper proposes a noval and enhanced image matching propagation algorithm which is applicable to building scenes.Adaptive support weight criterion was applied to decrease incorrect matches,and affine propagation was used to obtain best neighbor window of new matches.Experiments demonstrate that the proposed algorithm is efficient which can obtain large number of correct matches on building images.
作者
陈占军
王耀武
龚咏喜
Chen Zhanjun;Wang Yaowu;Gong Yongxi(School of Architecture,Harbin Institute of Technology,Shenzhen,Shenzhen 518055,Guangdong,China;Shenzhen Key Laboratory of Urban Planning and Decision Making,Harbin Institute of Technology,Shenzhen,Shenzhen 518055,Guangdong,China)
出处
《计算机应用与软件》
北大核心
2019年第12期214-219,231,共7页
Computer Applications and Software
基金
深圳市教育科学“十三五”规划2018年度课题(ybzz18011)
关键词
建筑物图像
自适应支持加权
仿射传递
图像匹配
匹配扩散
Building images
Adaptive support weight
Affine propagation
Image matching
Match propagation
