郑亦然,程 健.基于多视图三维重建方法在实验室系统中的应用研究[J].测控技术,2021,40(1):22-26
基于多视图三维重建方法在实验室系统中的应用研究
Application of Multi-View 3D Reconstruction Method in Laboratory System
  
DOI:10.19708/j.ckjs.2020.08.300
中文关键词:  多视图三维重建  SIFT  CMVS/PMVS  实验室系统
英文关键词:multi-view 3D reconstruction  SIFT  CMVS/PMVS  laboratory system
基金项目:
作者单位
郑亦然 中国科学技术大学 信息科学技术学院 
程 健 中国科学技术大学 信息科学技术学院 
摘要点击次数: 339
全文下载次数: 104
中文摘要:
      针对愈渐复杂的实验室系统,采用基于多视图的三维重建方法,实现对实验室系统中设备的结构重建。通过尺度不变特征变换(SIFT)匹配算法进行特征提取和匹配,并利用随机抽样一致性算法(RANSAC)和交叉过滤法降低误匹配率;通过多视图聚簇/基于面片的三维多视角立体视觉算法(CMVS/PMVS)获取稠密点云;利用PCL泊松曲面重构算法将稠密点云生成多边形网格,得到最终的重建图像。采用真实的实验室设备进行了验证和测试,实验结果表明,该方法可以有效快速地进行实验室设备的三维重建。
英文摘要:
      In view of the increasingly complex laboratory system,a multi-view-based three-dimensional (3D) reconstruction method is adopted to achieve the structural reconstruction of the laboratory equipment.First,the scale-invariant feature transform (SIFT) matching algorithm was used to extract and match the features,while the random sample consensus (RANSAC) and cross-filtering methods were used to reduce the false matching rate.Then,the dense point cloud was obtained through clustering multi-view stereo/patch-based multi-view stereo(CMVS/PMVS),which was then processed by PCL Poisson surface reconstruction algorithm to generate a polygonal mesh to obtain the final reconstructed image.Finally,it was verified and tested by real laboratory equipment.Experimental results show that this multi-view based method can perform the 3D reconstruction of laboratory equipments effectively and timely.
查看全文  查看/发表评论  下载PDF阅读器
关闭