Machine Vision and Applications

, Volume 8, Issue 3, pp 179–186

Range measurement from defocus gradient



Images taken of 3D objects are generally defocused. Depth-from-focus techniques use the defocus information to determine range. However, quantitative measurement of focus is difficult and requires accurate modeling of the point-spread function (PSF). We describe a new method that determines depth using the symmetry and smoothness of focus gradient with respect to the focus position. The technique is passive and uses a monocular imaging system. The performance for estimating range is experimentally demonstrated.

Key words

Depth from focus Shape from defocus Range sensing 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. Cardillo J, Sid-Ahmed MA (1991) 3-D position sensing using a passive monocular vision system. IEEE Trans Patt Anal Machine Intell 13:809–813Google Scholar
  2. Darrell T, Wohn K (1988) Pyramid based depth from focus. Proceedings IEEE Computer Vision and Pattern Recognition '88, pp 504–509Google Scholar
  3. Das S, Ahuja N (1989) Integrating multiresolution image acquisition and coarse-to-fine surface reconstruction. Proceedings of the Workshop on Interpretation of 3D Scenes, IEEE, 9–15Google Scholar
  4. Engelhardt K (1988) Acquisition of 3-D data by focus sensing. Applied Optics 27:4684–4689Google Scholar
  5. Ens J, Lawrence P (1993) An investigation of methods for determining depth from focus. IEEE Trans Patt Anal Machine Intell 15:97–108Google Scholar
  6. Grossman P (1987) Depth from focus. Patt Recogn Lett 5:63–69Google Scholar
  7. Horn BKP (1968) Focusing. MIT Project Mac. AI Memo No. 160, Massachusetts Institute of TechnologyGoogle Scholar
  8. Hwang T, Clark JJ, Yuille AL (1989) A depth recovery algorithm using defocus information. Proceedings IEEE Computer Vision and Pattern Recognition '89, 476–481Google Scholar
  9. Jarvis RA (1976) Focus optimization criteria for computer image processing. Microscope 24:163–180Google Scholar
  10. Kingslake R (1983) Optical system design. Academic Press, New YorkGoogle Scholar
  11. Krotkov E (1987) Focusing. Int J Comput Vision 1:223–237Google Scholar
  12. Lai S, Fu C, Chang S (1992) A generalized depth estimation algorithm with a single image. IEEE Trans Patt Anal Machine Intell 14:405–411Google Scholar
  13. Nayar SK (1992) Shape from focus system for rough surfaces. Proceedings of the Image Understanding Workshop, pp 593–606Google Scholar
  14. Pentland AP (1987) A new sense for depth of field. IEEE Trans Patt Anal Machine Intell 9:523–531Google Scholar
  15. Pentland AP, Darrell T, Turk M, Huang W (1989) A simple, real-time range camera. Proceedings of IEEE Computer Vision and Pattern Recognition '89, pp 256–261Google Scholar
  16. Rioux M, Biais F (1986) Compact three-dimensional camera for robotic applications. J Opt Soc Am A, 3:1518–1521Google Scholar
  17. Subbarao M (1987) Direct recovery of depth-map. Proceedings of the IEEE Comput Soc Workshop Comput Vision pp 58–65Google Scholar
  18. Subbarao M (1988) Parallel depth recovery by changing camera parameters. Proceedings of the IEEE 2nd International Conference on Computer Vision, pp 149–155Google Scholar
  19. Subbarao M (1989) Efficient depth recovery through inverse optics. In: H. Freeman (ed) Machine vision for inspection and measurement. Academic Press, New York, pp 101–126Google Scholar
  20. Subbarao M, Natarajan G (1988) Depth recovery from blurred edges. Proceedings of the IEEE Computer Vision and Pattern Recognition'88, pp 498–503Google Scholar

Copyright information

© Springer-Verlag 1995

Authors and Affiliations

  1. 1.Imaging Research Inc.Brock UniversitySt. CatharinesCanada
  2. 2.Department of Electrical and Computer EngineeringMcMaster UniversityHamiltonCanada
  3. 3.Department of Computer ScienceCurtin University of TechnologyPerthAustralia

Personalised recommendations