|
CVGResearch
Home |
3D Reconstruction We live in a three-dimensional world (with the elapsing time) and possess stereo vision thanks to the pair of our eyes (and our brain). Inspired by this natural endowment, a central problem of computer vision is to obtain 3D geometric shape information of the objects from planar images. This process is traditionally termed 3D reconstruction. 3D reconstruction of the objects and environments is useful in itself as it amounts to obtaining their 3D geometric measurement, and is also a helpful means for recognition problems since it eliminates the headache-causing variances produced by the perspective imaging process and variable illumination. |
|||||
| Relative Epipolar Motion of Tracked Features for Correspondence in Binocular Stereo. Hao Du, Danping Zou, Yan Qiu Chen. International Conference on Computer Vision, 2007. | ||||||
| Abstract. Most 3D reconstruction solutions focus on surfaces, and there has not been much research attention paid to the problem of reconstructing 3D scenes made up of large numbers of particles, while the ability to reconstruct such dynamic scenes is potentially very useful in many areas such as colony behavior research and visual modeling. This paper proposes an approach - Relative Epipolar Motion (REM) - towards solving the correspondence problem in stereopsis by utilizing the motion clue. It matches feature trajectories instead of the features themselves as used by existing methods. The proposed method has the following new capabilities: (1) It supports reconstructing dynamic 3D scenes of large number of undistinguishable drifting particles; (2) It is applicable to correspondence extablishment for dynamic surfaces made up of repetitive textures; (3) It offers an alternative way to project structured light in active mode for deforming surface reconstruction. Experiment results on both simulated and real-world scenes demonstrate its effectiveness. |
 |
|||||
|
|
||||||