Global depth from defocus with fixed camera parameters

Reconstruction depth from 2D images is an important research issue in computer vision, and depth from defocus (DFD) is an effective way which takes the blurred degree of the region images whose depth of field is limit as the tool of computing depth. Now though there are many DFD methods, they all need to change camera parameters in order to attain blurred images, such as the focal length of the lens, the radius of the lens. If cameras with high level of amplification are used, it is inhibitory to change camera parameters. Therefore, in this paper a novel DFD method is proposed. First, two different blurred images are captured through changing depth. Second, the blurred imaging model is constructed with the relative blurring and the diffusion equation, and the relation between depth and blurring is discussed from two aspects. Finally, the problem of computing depth is transformed into an optimization issue. The method proposed in this paper does not need to change camera parameters, so the process is very simple and can be used in some special applications. The simulation results show that this method can attain depth with high precision.

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

[2]  Stefano Soatto,et al.  Observing Shape from Defocused Images , 2004, International Journal of Computer Vision.

[3]  Naoki Asada,et al.  Edge and Depth from Focus , 2004, International Journal of Computer Vision.

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

[5]  Mats Gökstorp,et al.  Computing depth from out-of-focus blur using a local frequency representation , 1994, International Conference on Pattern Recognition.

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

[7]  Stefano Soatto,et al.  This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE IEEE TRANSACTION OF PATTERN RECO , 2022 .

[8]  Gaoliang Dai,et al.  Determining the residual nonlinearity of a high-precision heterodyne interferometer , 1999 .

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

[10]  Subhasis Chaudhuri,et al.  On defocus, diffusion and depth estimation , 2007, Pattern Recognit. Lett..

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

[12]  Shree K. Nayar,et al.  Real-Time Focus Range Sensor , 1996, IEEE Trans. Pattern Anal. Mach. Intell..

[13]  Stefano Soatto,et al.  Observing Shape from Defocused Images , 1999, Proceedings 10th International Conference on Image Analysis and Processing.

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

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

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

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

[18]  R. J. Shul,et al.  Depth and thermal stability of dry etch damage in GaN Schottky diodes , 1999 .