Adaptive Error Concealment of H.264/AVC Video Coding Standard for IPTV Application

Asheri, Hadi | 2011

1610 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 41819 (19)
  4. University: Sharif University of Technology
  5. Department: Computer Engineering
  6. Advisor(s): Rabiee, Hamid Reza
  7. Abstract:
  8. Error concealment is one of the effective ways to alleviate the effect of packet loss in video communication over error-prone environments. In order to estimate lost macro-blocks, we have employed Bayesian estimation as an efficient and robust framework. Gaussian process regression has been used as the modeling approach through this framework. Considering luminance component as Gaussian process,a minimum mean squared error estimation of the lost macro-block is obtained. This estimator, as a function of the existing data, is only determined by the covariance matrix defined over them. Therefore,the main step in Gaussian process regression, is construction of the convenient covariance matrix based on existing data. The essential answer to this requirement is usage of the kernel function which is a similarity measure between any two data points. Human visual system is very sensitive to edge distortion. Hence, we have incorporated the edge related information in the definition of the kernel function, providing an adaptive error concealment solution. Experimental results verify the ability of the proposed method in preserving edge information
    and producing smooth estimations of the lost macro-blocks. In this work, in order to recover lost motion vectors, we have also provided a Gaussian process regression solution based on the existing correctly received motion vectors in the neighboring area.Of the various video coding standards, H.264/AVC is the most widely developed by IPTV service providers. Experimenrtal results have have demonstrated that the proposed method is superior to the one employed by the H.264/AVC and comparable to the state-of-the-art methods in both temporal and spatial domains. Finally, we have concluded that stationary kernel functions are effective in smooth images and videos
  9. Keywords:
  10. H.264 Video Codec ; Spatial Error Concealment ; Bayesian Estimation ; Gaussian Process Regression Analysis ; Motion Vector Recovery

 Digital Object List

  • محتواي پايان نامه
  •   view


...see more