Loading...

Novel View Synthesis Using Light Field Camera

Bemana, Mojtaba | 2016

404 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 48896 (05)
  4. University: Sharif University of Technology
  5. Department: Electrical Engineering
  6. Advisor(s): Amini, Arash
  7. Abstract:
  8. In the last two decades, novel view synthesis has become increasingly important in the movie and game industries. The new generated views correspond to virtual viewpoints, where no real camera has taken snapshot. Depending on the amount of available information from the scene, different approaches have been proposed to create novel views. If the depth information is available, then by knowing camera relative motion, novel views can be constructed. The conventional cameras only record the intensity of the captured scene, and discard the depth information. In the recent concept of light field cameras, by locating an array of micro-lenses between the camera sensor and the main lens, one can capture not only a 2D image, but also the directions of incoming light rays. The additional information provided by the rays’ direction enables us to refocus images digitally and to estimate the scene depth. In 2012, the advent of first commercial light field camera – named Lytro™ – revolutionized the photography industry. Considering that Lytro output image contains both depth and intensity information, many recent works are dedicated to the area of depth estimation using light field cameras. The scope of this project is to synthesize a novel view of the scene using the images captured by the Lytro camera. In order to achieve this purpose, two images are taken from different viewpoints. Then utilizing one of the existing state-of-the-art depth estimation methods, the scene depth will be extracted. In the next step, the transformation between the images, which is a combination of rotation and translation, is estimated. In this thesis, various methods are proposed to improve both depth and camera motion estimations. Finally, knowing the relative motion and the depth map, a novel view of the scene is constructed
  9. Keywords:
  10. Novel View Synthesis ; Light Field Camera ; Relative Camera Motion Estimotion ; Depth Estimation

 Digital Object List

 Bookmark

No TOC