Depth from Defocus vs. Stereo: How Different Really Are They?

Depth from Focus (DFF) and Depth from Defocus (DFD) methods are theoretically unified with the geometric triangulation principle. Fundamentally, the depth sensitivities of DFF and DFD are not different than those of stereo (or motion) based systems having the same physical dimensions. Contrary to common belief, DFD does not inherently avoid the matching (correspondence) problem. Basically, DFD and DFF do not avoid the occlusion problem any more than triangulation techniques, but they are more stable in the presence of such disruptions. The fundamental advantage of DFF and DFD methods is the two-dimensionality of the aperture, allowing more robust estimation. We analyze the effect of noise in different spatial frequencies, and derive the optimal changes of the focus settings in DFD. These results elucidate the limitations of methods based on depth of field and provide a foundation for fair performance comparison between DFF/DFD and shape from stereo (or motion) algorithms.

[1]  Jianping Hu,et al.  A Wavelet Multiresolution Edge Analysis Method for Recovery of Depth from Defocused Images , 2001, WAA.

[2]  Shree K. Nayar,et al.  Microscopic shape from focus using active illumination , 1994, Proceedings of 12th International Conference on Pattern Recognition.

[3]  Alex Pentland,et al.  A New Sense for Depth of Field , 1985, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[4]  Shree K. Nayar,et al.  Real-time focus range sensor , 1995, Proceedings of IEEE International Conference on Computer Vision.

[5]  Edward H. Adelson,et al.  Single Lens Stereo with a Plenoptic Camera , 1992, IEEE Trans. Pattern Anal. Mach. Intell..

[6]  J. Koenderink,et al.  Photometric Invariants Related to Solid Shape , 1980 .

[7]  Worthy N. Martin,et al.  Image Motion Estimation From Motion Smear-A New Computational Model , 1996, IEEE Trans. Pattern Anal. Mach. Intell..

[8]  Charles V. Stewart,et al.  New Results in Automatic Focusing and a New Method for Combining Focus and Stereo , 1990, Other Conferences.

[9]  A A Friesem,et al.  Holographic elements with high efficiency and low aberrations for helmet displays. , 1989, Applied optics.

[10]  Murali Subbarao,et al.  Accurate reconstruction of three-dimensional shape and focused image from a sequence of noisy defocused images , 1997, Other Conferences.

[11]  H.N. Nair,et al.  Robust focus ranging , 1992, Proceedings 1992 IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[12]  Peter Lawrence,et al.  An Investigation of Methods for Determining Depth from Focus , 1993, IEEE Trans. Pattern Anal. Mach. Intell..

[13]  Trevor Darrell,et al.  Pyramid based depth from focus , 1988, Proceedings CVPR '88: The Computer Society Conference on Computer Vision and Pattern Recognition.

[14]  Murali Subbarao,et al.  Depth from defocus and rapid autofocusing: a practical approach , 1992, Proceedings 1992 IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[15]  A. Pentland.  Finding the illuminant direction , 1982 .

[16]  Takakazu Ishimatsu,et al.  Range finder using circular dynamic stereo , 1998, Proceedings. Fourteenth International Conference on Pattern Recognition (Cat. No.98EX170).

[17]  John Oliensis,et al.  Uniqueness in shape from shading , 1991, International Journal of Computer Vision.

[18]  Bernd Girod,et al.  Depth from Defocus of Structured Light , 1990, Other Conferences.

[19]  Akbar R. Saadat,et al.  Simple, general, and mathematically tractable way to sense depth in a single image , 1995, Optics & Photonics.

[20]  Subhasis Chaudhuri,et al.  Optimal selection of camera parameters for recovery of depth from defocused images , 1997, Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[21]  Yoav Y. Schechner,et al.  The optimal axial interval in estimating depth from defocus , 1999, Proceedings of the Seventh IEEE International Conference on Computer Vision.

[22]  Murali Subbarao.  Parallel Depth Recovery By Changing Camera Parameters , 1988, [1988 Proceedings] Second International Conference on Computer Vision.

[23]  Paul J. Besl,et al.  Active, optical range imaging sensors , 1988, Machine Vision and Applications.

[24]  Rama Chellappa,et al.  Estimation of Illuminant Direction, Albedo, and Shape from Shading , 1991, IEEE Trans. Pattern Anal. Mach. Intell..

[25]  Shmuel Peleg,et al.  A Three-Frame Algorithm for Estimating Two-Component Image Motion , 1992, IEEE Trans. Pattern Anal. Mach. Intell..

[26]  Murali Subbarao,et al.  Integration of defocus and focus analysis with stereo for 3D shape recovery , 1997, Other Conferences.

[27]  Eero P. Simoncelli,et al.  Direct Differential Range Estimation Using Optical Masks , 1996, ECCV.

[28]  Ray A. Jarvis,et al.  A Perspective on Range Finding Techniques for Computer Vision , 1983, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[29]  Naoki Asada,et al.  Seeing Behind the Scene: Analysis of Photometric Properties of Occluding Edges by the Reversed Projection Blurring Model , 1998, IEEE Trans. Pattern Anal. Mach. Intell..

[30]  Alex Pentland,et al.  Simple range cameras based on focal error , 1994 .

[31]  Alan L. Yuille,et al.  A depth recovery algorithm using defocus information , 1989, Proceedings CVPR '89: IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[32]  Wojciech Chojnacki,et al.  Can the Sun's Direction be Estimated from an Image Prior to the Computation of Object Shape? , 2004, Journal of Mathematical Imaging and Vision.

[33]  Alfred M. Bruckstein,et al.  Tracking Level Sets by Level Sets: A Method for Solving the Shape from Shading Problem , 1995, Comput. Vis. Image Underst..

[34]  H. H. Hopkins.  The frequency response of a defocused optical system , 1955, Proceedings of the Royal Society of London. Series A. Mathematical and Physical Sciences.

[35]  V. Bove.  Entropy-based depth from focus , 1993 .

[36]  G Häusler,et al.  Acquisition of 3-D data by focus sensing. , 1988, Applied optics.

[37]  Richard Szeliski,et al.  Impossible Shaded Images , 1993, IEEE Trans. Pattern Anal. Mach. Intell..

[38]  Wilson S. Geisler,et al.  Maximum-likelihood depth-from-defocus for active vision , 1995, Proceedings 1995 IEEE/RSJ International Conference on Intelligent Robots and Systems. Human Robot Interaction and Cooperative Robots.

[39]  Joel S. Fox.  Range from translational motion blurring , 1988, Proceedings CVPR '88: The Computer Society Conference on Computer Vision and Pattern Recognition.

[40]  Ahmed M. Darwish.  3D from focus and light stripes , 1994, Other Conferences.

[41]  Takashi Matsuyama,et al.  Depth measurement by the multi-focus camera , 1998, Proceedings. 1998 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (Cat. No.98CB36231).

[42]  John Oliensis,et al.  A global algorithm for shape from shading , 1993, 1993 (4th) International Conference on Computer Vision.

[43]  Bernard Heit,et al.  Monocular depth perception by evaluation of the blur in defocused images , 1994, Proceedings of 1st International Conference on Image Processing.

[44]  Eero P. Simoncelli,et al.  Range estimation by optical differentiation. , 1998, Journal of the Optical Society of America. A, Optics, image science, and vision.

[45]  A. Lynn Abbott,et al.  Active Stereo: Integrating Disparity, Vergence, Focus, Aperture and Calibration for Surface Estimation , 1993, IEEE Trans. Pattern Anal. Mach. Intell..

[46]  Michal Irani,et al.  Computing occluding and transparent motions , 1994, International Journal of Computer Vision.

[47]  A. Lynn Abbott,et al.  Surface Reconstruction By Dynamic Integration Of Focus, Camera Vergence, And Stereo , 1988, [1988 Proceedings] Second International Conference on Computer Vision.

[48]  C. Swain,et al.  Depth estimation from image defocus using fuzzy logic , 1994, Proceedings of 1994 IEEE 3rd International Fuzzy Systems Conference.

[49]  Alex Pentland.  Linear shape from shading , 2004, International Journal of Computer Vision.

[50]  Alfred M. Bruckstein,et al.  Global Shape from Shading , 1996, Comput. Vis. Image Underst..

[51]  Mohan M. Trivedi,et al.  Active vision system for depth extraction using multiprimitive hierarchical stereo analysis and multiple depth cues , 1993, Defense, Security, and Sensing.

[52]  Ruzena Bajcsy,et al.  Active vision for reliable ranging: Cooperating focus, stereo, and vergence , 1993, International Journal of Computer Vision.

[53]  Hsien-Che Lee.  Review of image-blur models in a photographic system using the principles of optics , 1990 .

[54]  Murali Subbarao,et al.  Depth from defocus: A spatial domain approach , 1994, International Journal of Computer Vision.

[55]  W. Cathey,et al.  Defocus transfer function for circularly symmetric pupils. , 1997, Applied optics.

[56]  A.L Yuille,et al.  Zero crossings on lines of curvature , 1989, Comput. Vis. Graph. Image Process..

[57]  Shree K. Nayar,et al.  Shape from focus system , 1992, Proceedings 1992 IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[58]  Henrik I. Christensen,et al.  Cooperative depth extraction , 1993 .

[59]  Shree K. Nayar,et al.  Minimal operator set for passive depth from defocus , 1996, Proceedings CVPR IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[60]  M. Turk,et al.  A simple, real-time range camera , 1989, Proceedings CVPR '89: IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[61]  Murali Subbarao,et al.  Depth from defocus by changing camera aperture: a spatial domain approach , 1993, Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.

[62]  Subhasis Chaudhuri,et al.  A block shift-variant blur model for recovering depth from defocused images , 1995, Proceedings., International Conference on Image Processing.

[63]  Ronen Basri,et al.  Separation of Transparent Layers using Focus , 1998, Sixth International Conference on Computer Vision (IEEE Cat. No.98CH36271).

[64]  Takayuki Okatani,et al.  On the Classification of Singular Points for the Global Shape from Shading Problem: A Study of the Constraints Imposed by Isophotes , 1998, ACCV.

[65]  Francois Blais,et al.  Compact three-dimensional camera for robotic applications , 1986 .

[66]  C. A. Burbeck,et al.  Occlusion edge blur: a cue to relative visual depth. , 1996, Journal of the Optical Society of America. A, Optics, image science, and vision.

[67]  Steven A. Shafer,et al.  Depth from focusing and defocusing , 1993, Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.