High Efficiency Video Coding via Predictive Texture Synthesis
碩士 === 國立中山大學 === 電機工程學系研究所 === 101 === In recent years, video coding technology such as H.264/AVC and High Efficiency Video Coding (HEVC) has matured. Among the coding methods, intra coding is a very important one. The prediction accuracy of intra coding is the key to performance enhancement. This...
Main Authors: | , |
---|---|
Other Authors: | |
Format: | Others |
Language: | en_US |
Published: |
2013
|
Online Access: | http://ndltd.ncl.edu.tw/handle/6h26jd |
id |
ndltd-TW-101NSYS5442127 |
---|---|
record_format |
oai_dc |
spelling |
ndltd-TW-101NSYS54421272019-05-15T21:02:52Z http://ndltd.ncl.edu.tw/handle/6h26jd High Efficiency Video Coding via Predictive Texture Synthesis 基於預測紋理合成之高效能視訊編碼 Tsungyi-Yi Tseng 曾宗益 碩士 國立中山大學 電機工程學系研究所 101 In recent years, video coding technology such as H.264/AVC and High Efficiency Video Coding (HEVC) has matured. Among the coding methods, intra coding is a very important one. The prediction accuracy of intra coding is the key to performance enhancement. This thesis presents a high efficiency video coding method using predictive texture synthesis. First, the proposed method selects two predictive blocks with different prediction directions and synthesizes these two prediction blocks. The proposed method then employs two core methods (combined mode selection method and weight recover technique) to enhance coding performance of intra coding. The combined mode selection method reduces the bits required to suggest which two modes are selected in the decoder and the computational complexity of selecting the two modes; meanwhile, the weight recover technique saves the bits required for weights in the decoder. Experimental results show that the proposed method improves the coding performance of intra prediction. Chia-Hung Yeh 葉家宏 2013 學位論文 ; thesis 63 en_US |
collection |
NDLTD |
language |
en_US |
format |
Others
|
sources |
NDLTD |
description |
碩士 === 國立中山大學 === 電機工程學系研究所 === 101 === In recent years, video coding technology such as H.264/AVC and High Efficiency Video Coding (HEVC) has matured. Among the coding methods, intra coding is a very important one. The prediction accuracy of intra coding is the key to performance enhancement. This thesis presents a high efficiency video coding method using predictive texture synthesis. First, the proposed method selects two predictive blocks with different prediction directions and synthesizes these two prediction blocks. The proposed method then employs two core methods (combined mode selection method and weight recover technique) to enhance coding performance of intra coding. The combined mode selection method reduces the bits required to suggest which two modes are selected in the decoder and the computational complexity of selecting the two modes; meanwhile, the weight recover technique saves the bits required for weights in the decoder. Experimental results show that the proposed method improves the coding performance of intra prediction.
|
author2 |
Chia-Hung Yeh |
author_facet |
Chia-Hung Yeh Tsungyi-Yi Tseng 曾宗益 |
author |
Tsungyi-Yi Tseng 曾宗益 |
spellingShingle |
Tsungyi-Yi Tseng 曾宗益 High Efficiency Video Coding via Predictive Texture Synthesis |
author_sort |
Tsungyi-Yi Tseng |
title |
High Efficiency Video Coding via Predictive Texture Synthesis |
title_short |
High Efficiency Video Coding via Predictive Texture Synthesis |
title_full |
High Efficiency Video Coding via Predictive Texture Synthesis |
title_fullStr |
High Efficiency Video Coding via Predictive Texture Synthesis |
title_full_unstemmed |
High Efficiency Video Coding via Predictive Texture Synthesis |
title_sort |
high efficiency video coding via predictive texture synthesis |
publishDate |
2013 |
url |
http://ndltd.ncl.edu.tw/handle/6h26jd |
work_keys_str_mv |
AT tsungyiyitseng highefficiencyvideocodingviapredictivetexturesynthesis AT céngzōngyì highefficiencyvideocodingviapredictivetexturesynthesis AT tsungyiyitseng jīyúyùcèwénlǐhéchéngzhīgāoxiàonéngshìxùnbiānmǎ AT céngzōngyì jīyúyùcèwénlǐhéchéngzhīgāoxiàonéngshìxùnbiānmǎ |
_version_ |
1719108595898908672 |