New computer vision techniques use photo datasetto rapidly build detailed 3D models. Computer-vision researchers have explored many approachesto city-scale 3D reconstruc on. Amongthese systems stands out VisualSfM developed bythe University of Washington & Google Inc. It is aopen source GUI applica on of a Structure fromMo on (SfM) so! ware that uses a feature extractorcalled Si! GPU and the Mul core BundleAdjustment. In addi on it embeds the CMVS/PMVS2 able to reconstruct dense 3D point cloud.Our goal is to demonstrate the metric accuracy ofVisualSfM+CMVS/PMVS2 and that to get run it,you can use an unstructured photo dataset butthe result improves if you use a structured photodataset. The approach has been tested on severallarge datasets with structured images.
|Numero di pagine||8|
|Stato di pubblicazione||Published - 2013|