Postprocessing of images coded using block DCT at low bit rates.

Sun, Deqing.Thesis (M.Phil.)--Chinese University of Hong Kong, 2007.Includes bibliographical references (leaves 86-91).Abstracts in English and Chinese.Abstract --- p.i摘要 --- p.iiiContributions --- p.ivAcknowledgement --- p.viAbbreviations --- p.xviiiNotations --- p.xxiChapter 1 --- Introduction --- p.1Chapter 1.1 --- Image compression and postprocessing --- p.1Chapter 1.2 --- A brief review of postprocessing --- p.3Chapter 1.3 --- Objective and methodology of the research --- p.7Chapter 1.4 --- Thesis organization --- p.8Chapter 1.5 --- A note on publication --- p.9Chapter 2 --- Background Study --- p.11Chapter 2.1 --- Image models --- p.11Chapter 2.1.1 --- Minimum edge difference (MED) criterion for block boundaries --- p.12Chapter 2.1.2 --- van Beek's edge model for an edge --- p.15Chapter 2.1.3 --- Fields of experts (FoE) for an image --- p.16Chapter 2.2 --- Degradation models --- p.20Chapter 2.2.1 --- Quantization constraint set (QCS) and uniform noise --- p.21Chapter 2.2.2 --- Narrow quantization constraint set (NQCS) --- p.22Chapter 2.2.3 --- Gaussian noise --- p.22Chapter 2.2.4 --- Edge width enlargement after quantization --- p.25Chapter 2.3 --- Use of these models for postprocessing --- p.27Chapter 2.3.1 --- MED and edge models --- p.27Chapter 2.3.2 --- The FoE prior model --- p.27Chapter 3 --- Postprocessing using MED and edge models --- p.28Chapter 3.1 --- Blocking artifacts suppression by coefficient restoration --- p.29Chapter 3.1.1 --- AC coefficient restoration by MED --- p.29Chapter 3.1.2 --- General derivation --- p.31Chapter 3.2 --- Detailed algorithm --- p.34Chapter 3.2.1 --- Edge identification --- p.36Chapter 3.2.2 --- Region classification --- p.36Chapter 3.2.3 --- Edge reconstruction --- p.37Chapter 3.2.4 --- Image reconstruction --- p.37Chapter 3.3 --- Experimental results --- p.38Chapter 3.3.1 --- Results of the proposed method --- p.38Chapter 3.3.2 --- Comparison with one wavelet-based method --- p.39Chapter 3.4 --- On the global minimum of the edge difference . . --- p.41Chapter 3.4.1 --- The constrained minimization problem . . --- p.41Chapter 3.4.2 --- Experimental examination --- p.42Chapter 3.4.3 --- Discussions --- p.43Chapter 3.5 --- Conclusions --- p.44Chapter 4 --- Postprocessing by the MAP criterion using FoE --- p.49Chapter 4.1 --- The proposed method --- p.49Chapter 4.1.1 --- The MAP criterion --- p.49Chapter 4.1.2 --- The optimization problem --- p.51Chapter 4.2 --- Experimental results --- p.52Chapter 4.2.1 --- Setting algorithm parameters --- p.53Chapter 4.2.2 --- Results --- p.56Chapter 4.3 --- Investigation on the quantization noise model . . --- p.58Chapter 4.4 --- Conclusions --- p.61Chapter 5 --- Conclusion --- p.71Chapter 5.1 --- Contributions --- p.71Chapter 5.1.1 --- Extension of the DCCR algorithm --- p.71Chapter 5.1.2 --- Examination of the MED criterion --- p.72Chapter 5.1.3 --- Use of the FoE prior in postprocessing . . --- p.72Chapter 5.1.4 --- Investigation on the quantization noise model --- p.73Chapter 5.2 --- Future work --- p.73Chapter 5.2.1 --- Degradation model --- p.73Chapter 5.2.2 --- Efficient implementation of the MAP method --- p.74Chapter 5.2.3 --- Postprocessing of compressed video --- p.75Chapter A --- Detailed derivation of coefficient restoration --- p.76Chapter B --- Implementation details of the FoE prior --- p.81Chapter B.1 --- The FoE prior model --- p.81Chapter B.2 --- Energy function and its gradient --- p.83Chapter B.3 --- Conjugate gradient descent method --- p.84Bibliography --- p.8