Home | Publications | Stanford Range Image Data | Monocular Depth Estimation | Single Image 3D Reconstruction

3-D Reconstruction from Sparse Views using Monocular Vision

We consider the task of creating a 3-d model of a large novel environment, given only a small number of images of the scene. See more details on algorithm here. This page shows some examples of larger 3d models created using Monocular Vision (July 2007).

For viewing videos, we suggest: Divx for Windows, For Linux try Mplayer. For viewing VRMLs, we suggest Cortona plugin for Internet explorer in Windows. Browser plugin takes less than a minute to install.

Example 1 Example 2 Example 3 Example 4

Example 1: Images with small overlap

Original Images:

Models: Link to model



Example 2: Images with repeated structures


Original Images

Models Link to model



Example 3: 3-D model with 3 images at sparse views

Images

Models Video Link to model



Example 4: Model of a big area using only 4 images


Images

Models: Video, Link to model