3D reconstruction implementation using OpenCV.
Key points are matched between images using a Lucas-Kanade tracker. The matched keypoints are used in a RANSAC algorithm to find camera matrices, and 3D positions are reconstructed. Delaunay triangulation creates a 3D mesh from the point cloud.
(Data from robots.ox.ac.uk/~vgg/data/data-mview.html)
Loading more stuff…
Hmm…it looks like things are taking a while to load. Try again?