Paper Details: Downloads: 2405
Serial Number: P1151811613
Title: Bayesian Early Mode Decision Technique for Multiview Video Coding
Authors: Aisha Gul and Shadan Khattak
Abstract: Multi-view Video Coding (MVC) is a technique which is used for the efficient compression of Multi-view videos (MVV). A problem with MVC is its huge computational complexity. This paper studies the application of Bayesian decision theory to reduce the encoding complexity of MVC. Moreover, it proposes to integrate a Bayesian decision based early mode decision based technique and an adaptive disparity estimation technique in such a way that the gains of both the techniques add up to reduce the encoding complexity of MVC more effectively compared to each of these approaches applied separately. The proposed combined approach results in saving 66.53% of encoding time at the cost of only 0.03 dB in peak signal-to-noise ratio (PSNR) and 2.13% increase in bitrate compared to the standard MVC implementation (JMVC 8.5).
Keywords: Video compression, Compression standards, H.264, Multiview video coding.
Journal/Conference: International Journal of Graphics, Vision and Image Processing
Volume: 18
Issue: 1
Submission Date: 3/14/2018 12:00:00 AM
Review Date: 5/8/2018 12:00:00 AM
Publishing Date: 5/19/2018 12:00:00 AM
Article Downloads: 2405
Download:

Facebook