Fast automatic registration of range images from 3D imaging systems using sphere targets

Automation in Construction(2009)

引用 64|浏览11
暂无评分
摘要
The use of 3D imaging systems (e.g., laser scanners) in construction has grown significantly in the past decade. Range images acquired with such systems often require registration. This paper describes an automatic method to rapidly locate spheres and perform a registration based on three pairs of matching points (centers of fitted spheres) in two range images. The proposed method is directly applicable for regularly gridded datasets obtained with instruments that are typically used for construction applications and whose maximum ranges are greater than 50 m. A lab was scanned from two locations at three different scan densities. Four spheres were located in the lab, and the total number of points hitting the four spheres was a small fraction (<0.01%) of all the points in the dataset. At the highest scan density, the registration of two datasets with 6.4×106 and 3.4×106 points is obtained in less than 30 s. At the medium scan density, two range images with 1.6×106 and 0.8×106 points can be registered in less than 2 s.
更多
查看译文
关键词
3D imaging system,Automated object recognition,Sphere fitting,Target-based registration
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要