Journal: Journal of Imaging
Article Title: Guidelines for Accurate Multi-Temporal Model Registration of 3D Scanned Objects
doi: 10.3390/jimaging9020043
Figure Lengend Snippet: Registration error relative to the cross-sectional area of two selected features (planar surfaces) in terms of % field of view (% FOV). The small error when selecting block features accounting for 4% FOV is similar to the average error when using the entire 3D model (tibia + block; 96% FOV) to optimize the transformation.
Article Snippet: Landi et al. used a portable 3D laser scanner (FreeScan UE7; Geomagic Control X; maximum FOV = 2652 cm 2 ) to analyze the deformation of machine guards before and after ballistic penetration, where 3D models were registered using intrinsic features consisting of the undeformed edges of the guards, with registration errors of approximately 300 μm [ ].
Techniques: Blocking Assay, Transformation Assay