Back to results list
Please use this identifier to cite or link to this item:
|Title:||New coding technologies for videos||Authors:||Wang, Li Li||Keywords:||Digital video.
Hong Kong Polytechnic University -- Dissertations
|Issue Date:||2013||Publisher:||The Hong Kong Polytechnic University||Abstract:||In order to save transmission bandwidth and storage space, a series of video coding standards have been established during the past two decades, including MPEG-1/2/4, H.261/H.263 and the current H.264/AVC standard. As compared with earlier standards, H.264/AVC has achieved a significant improvement in the rate-distortion performance. That is, it is possible to save bit rate up to 50% of MPEG-2, H.263, or MPEG-4 Part 2 while keeping the video quality about the same. However, the complexity of an encoder based on H.264/AVC standard is thus very high. Therefore, in this thesis, some novel techniques are developed to speed up an encoding process of inter and intra coding in H.264/AVC standard in the initial part of our research work. Experimental results show that the proposed algorithms can achieve a remarkable reduction of encoding complexity with a negligible loss in rate distortion performance compared with those in H.264/AVC. When compared with other algorithms in the literature, the proposed algorithms also give the best performance in terms of both coding efficiency and coding complexity. With the development of video coding techniques, high definition and ultra-high definition videos are gradually entering our daily life. The demands for the fidelity of the coded videos are higher. The current H.264/AVC standard will soon be unable to efficiently meet the compression demand for transmission or storage at much low bit rate. To develop some new video coding techniques which can give a better rate distortion performance than H.264/AVC standard is increasingly important. Hence this thesis also develops two efficient hierarchical intra prediction algorithms for videos based on H.264/AVC standard. Experimental results show that, as compared to the conventional methods, the proposed algorithm can significantly enhance the compression efficiency of lossless intra coding in H.264/AVC.
However, the H.264/AVC standard may not be still the best in terms of the compression of such larger and higher quality video contents. To overcome this problem, a new generation of video coding standard, which is called High Efficiency Video Coding (VCEG) standard, is being developed by the Joint Collaborative Team-Video Coding (JCT-VC). The objective of HEVC is to achieve a 50% reduction in bitrate compared with the high profile of H.264/AVC standard at the expense of computational complexity but without degrading the image quality. It will be ratified as an international standard in early 2013. Compared with H.264/AVC, many advanced coding tools have been adopted in the HEVC standard. For example, up to 35 modes are available for intra prediction of each prediction unit in HEVC. This can provide more accurate predictions and thereby improve the compression efficiency of intra coding. However, the complexity of an encoder based on the HEVC standard is thereby extremely high. We have found that there is still room for improvement in terms of high intra compression efficiency with less encoding complexity. In the last part of this thesis, we propose an adaptive intra modes skipping algorithm for mode decision and signaling processes for HEVC. The proposed algorithm not only reduces the encoding complexity. Importantly, it also enhances the compression efficiency of intra coding in the HEVC standard.
|Description:||xix, 177 p. : ill. (some col.) ; 30 cm.
PolyU Library Call No.: [THS] LG51 .H577P EIE 2013 Wang
|URI:||http://hdl.handle.net/10397/6193||Rights:||All rights reserved.|
|Appears in Collections:||Thesis|
Show full item record
Files in This Item:
|b26160614_link.htm||For PolyU Users||203 B||HTML||View/Open|
|b26160614_ir.pdf||For All Users (Non-printable)||3.3 MB||Adobe PDF||View/Open|
Citations as of Sep 18, 2018
Citations as of Sep 18, 2018
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.