• DocumentCode
    3327407
  • Title

    The refocusing distance of a standard plenoptic photograph

  • Author

    Hahne, Christopher ; Aggoun, Amar ; Velisavljevic, Vladan

  • Author_Institution
    Dept. of Comput. Sci. & Technol., Univ. of Bedfordshire, Luton, UK
  • fYear
    2015
  • fDate
    8-10 July 2015
  • Firstpage
    1
  • Lastpage
    4
  • Abstract
    In the past years, the plenoptic camera aroused an increasing interest in the field of computer vision. Its capability of capturing three-dimensional image data is achieved by an array of micro lenses placed in front of a traditional image sensor. The acquired light field data allows for the reconstruction of photographs focused at different depths. Given the plenoptic camera parameters, the metric distance of refocused objects may be retrieved with the aid of geometric ray tracing. Until now there was a lack of experimental results using real image data to prove this conceptual solution. With this paper, the very first experimental work is presented on the basis of a new ray tracing model approach, which considers more accurate micro image centre positions. To evaluate the developed method, the blur metric of objects in a refocused image stack is measured and compared with proposed predictions. The results suggest quite an accurate approximation for distant objects and deviations for objects closer to the camera device.
  • Keywords
    cameras; computer vision; image capture; image sensors; microlenses; optical images; ray tracing; computer vision; geometric ray tracing model; image sensor; microimage centre position; microlenses; plenoptic camera parameter; standard plenoptic photograph refocusing distance; three-dimensional image data capture; Cameras; Lenses; Measurement; Optical imaging; Optical sensors; Ray tracing; Standards; Light field; camera; distance estimate; plenoptic; refocusing;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    3DTV-Conference: The True Vision - Capture, Transmission and Display of 3D Video (3DTV-CON), 2015
  • Conference_Location
    Lisbon
  • Type

    conf

  • DOI
    10.1109/3DTV.2015.7169363
  • Filename
    7169363