Combining Shape-from-Motion Output with Partial Metadata | IEEE Conference Publication | IEEE Xplore