Depth From Distortions

Traditionally, images distortions have been regarded as a flaw that needed to be corrected. Much effort has been focused on “undistortion” methods to remedy this shortcoming. However, we see distortions as encoding crucial information about scene structure. In this paper we describe how scene depth is encoded in the distortions of images acquired with non-single viewpoint cameras or by mosaic construction. We present our framework to exploit these distortions for 3D euclidean reconstruction of scene features from a single image or view. We present methods specifically designed for features such as straight lines, circles or conics and show how it applies to general planar curves as well. Rigorous experimentation using simulations and synthetic images from catadioptric sensors and mosaics are presented.

[1]  M. Srinivasan,et al.  Reflective surfaces for panoramic imaging. , 1997, Applied optics.

[2]  Amnon Shashua,et al.  Trajectory Triangulation: 3D Reconstruction of Moving Points from a Monocular Image Sequence , 2000, IEEE Trans. Pattern Anal. Mach. Intell..

[3]  Shree K. Nayar,et al.  A theory of catadioptric image formation , 1998, Sixth International Conference on Computer Vision (IEEE Cat. No.98CH36271).

[4]  Vincenzo Caglioti,et al.  On the localization of straight lines in 3D space from single 2D images , 2005, 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05).

[5]  Alfred M. Bruckstein,et al.  Omniview cameras with curved surface mirrors , 2000, Proceedings IEEE Workshop on Omnidirectional Vision (Cat. No.PR00704).

[6]  Shmuel Peleg,et al.  Mosaicing on Adaptive Manifolds , 2000, IEEE Trans. Pattern Anal. Mach. Intell..

[7]  Seth J. Teller,et al.  Determining the Lines Through Four Lines , 1999, J. Graphics, GPU, & Game Tools.

[8]  Peter F. Sturm,et al.  A Generic Concept for Camera Calibration , 2004, ECCV.

[9]  R. Andrew Hicks,et al.  Geometric distributions for catadioptric sensor design , 2001, Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. CVPR 2001.

[10]  Ambjörn Naeve.  Structure from Translational Observer Motion , 1997, AFPAC.

[11]  Shree K. Nayar,et al.  A perspective on distortions , 2003, 2003 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2003. Proceedings..

[12]  Guillermo Sapiro,et al.  3D reconstruction from a single image , 2008 .

[13]  Shree K. Nayar,et al.  Non-perspective imaging systems , 2003 .

[14]  Shree K. Nayar,et al.  The Raxel Imaging Model and Ray-Based Calibration , 2005, International Journal of Computer Vision.

[15]  Yasushi Yagi,et al.  Real-time generation of environmental map and obstacle avoidance using omnidirectional image sensor with conic mirror , 1991, Proceedings. 1991 IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[16]  Shree K. Nayar,et al.  Non-Single Viewpoint Catadioptric Cameras: Geometry and Analysis , 2006, International Journal of Computer Vision.

[17]  Douglas Lanman,et al.  Reconstructing a 3D Line from a Single Catadioptric Image , 2006, Third International Symposium on 3D Data Processing, Visualization, and Transmission (3DPVT'06).

[18]  Czech Technical,et al.  Mirror Design for an Omnidirectional Camera with a Uniform Cylindrical Projection when Using the SVAVISCA Sensor , 2001 .

[19]  Tom,et al.  Central Panoramic Cameras: Design and Geometry Central Panoramic Cameras: Design and Geometry Tomm a S Svoboda, Tomm a S Pajdla and Vv Aclav Hlavv a C , 1998 .

[20]  Ian D. Reid,et al.  Single View Metrology , 2000, International Journal of Computer Vision.