HIGH SPEED AUTOMATIC DEPTH MAP GENERATION FOR 3D TELEVISION

  • Amisha H. Somaiya Vivekanand Education Society’s Institute of Technology, University of Mumbai, Maharashtra, India
  • Ramesh K. Kulkarni Vivekanand Education Society’s Institute of Technology, University of Mumbai, Maharashtra, India

Abstract

3D Television is the latest advancement in television viewing. Creating enough stereoscopic (S3D) material with stereoscopic cameras for 3DTV is time consuming & expensive. Thus the conversion of the vast collection of already existing 2D images/videos to S3D is essential. Monoscopic video content can be efficiently converted to stereo by using depth maps. The effectiveness of 2D to 3D video conversion depends on the accuracy of the generated depth maps. However, current techniques either use single monocular depth cues, which are restrained to a specific set of images/videos or combine multiple monocular cues, thus increasing the complexity and execution time of the system. The proposed algorithm alleviates these problems by using a novel depth map generation algorithm that can be used over a wide range of arbitrary videos. The high efficiency of this automatic algorithm without requiring any prior training combined with its high execution speed make it ideal for application in 3D television broadcasting industry.

Downloads

Download data is not yet available.
Published
2012-12-29
How to Cite
Somaiya, A. H., & Kulkarni, R. K. (2012). HIGH SPEED AUTOMATIC DEPTH MAP GENERATION FOR 3D TELEVISION. European Scientific Journal, ESJ, 8(30). https://doi.org/10.19044/esj.2012.v8n30p%p