Title: Constructing interactive multi-view videos based on image-based rendering
Authors: Shih-Ming Chang; Joseph C. Tsai; Shwu-Huey Yen; Timothy K. Shih
Addresses: Department of Computer Science and Information Engineering, Tamkang University, New Taipei City, 25137, Taiwan ' Department of Computer Science and Information Engineering, Tamkang University, New Taipei City, 25137, Taiwan ' Department of Computer Science and Information Engineering, Tamkang University, New Taipei City, 25137, Taiwan ' Department of Computer Science and Information Engineering, National Central University, Taoyuan County, 30201, Taiwan
Abstract: In this paper, we use image-based rendering (IBR) to develop a scene rotation mechanism. We shot several images in the same scene and computed the angles between images. A video is then composed, allowing users to select viewing angles when the video is playing. We made three kinds of assumptions that may affect the resulting video, and proved our assumptions by a series of experiments. Finally, we use video of realistic scenario and produce interactive video by the proposed method. The contribution also includes techniques to compute geometric parameters of the scene from one or more images.
Keywords: SIFT algorithm; mean-shift algorithm; homography matrix; image-based rendering; IBR; multi-view videos; interactive videos; scene rotation; viewing angles; viewing angle selection.
DOI: 10.1504/IJCSE.2015.070996
International Journal of Computational Science and Engineering, 2015 Vol.10 No.4, pp.402 - 414
Received: 01 Sep 2012
Accepted: 05 Sep 2012
Published online: 05 Aug 2015 *