結(jié)合率失真代價(jià)預(yù)測的HEVC快速CU劃分
發(fā)布時(shí)間：2019-01-21 08:37
【摘要】：針對高效率視頻編碼(high efficiency video coding,HEVC)標(biāo)準(zhǔn)編碼復(fù)雜度較高的問題,提出了一種快速編碼單元(coding unit,CU)劃分方法。首先,結(jié)合拉格朗日率失真優(yōu)化理論及相關(guān)實(shí)驗(yàn)數(shù)據(jù)分析得到相鄰CU深度對應(yīng)的失真及碼率分別滿足線性關(guān)系,并利用此關(guān)系,建立了率失真代價(jià)的預(yù)測模型。利用此模型,可以在編碼當(dāng)前CU深度后快速預(yù)測得到下一CU深度的率失真代價(jià),并最終通過代價(jià)比較,判斷是否需要繼續(xù)進(jìn)行CU劃分。實(shí)驗(yàn)結(jié)果表明,相比于HEVC測試模型HM12.0,針對低時(shí)延與隨機(jī)接入編碼結(jié)構(gòu),提出的方法的BD-rate分別增加了0.2%與0.6%,同時(shí),編碼時(shí)間分別減少了33.2%和38.9%。
[Abstract]:A fast coding unit (coding unit,CU) partition method is proposed to solve the problem of high complexity of high efficiency video coding (high efficiency video coding,HEVC) standard coding. Firstly, based on Lagrangian rate-distortion optimization theory and related experimental data analysis, the distortion and bit-rate corresponding to the adjacent CU depth are obtained to satisfy the linear relationship, and the prediction model of rate-distortion cost is established by using this relationship. By using this model, the rate-distortion cost of the next CU depth can be predicted quickly after encoding the current CU depth, and finally, the cost comparison can be used to determine whether it is necessary to continue the CU partitioning. The experimental results show that compared with the HEVC test model HM12.0, the BD-rate of the proposed method increases by 0.2% and 0.6% respectively for low delay and random access coding structures, and at the same time, The coding time was reduced by 33.2% and 38.9%, respectively.
【作者單位】： 西北工業(yè)大學(xué)電子信息學(xué)院;
【基金】：國家自然科學(xué)基金(61371089) 中央高校基本科研業(yè)務(wù)費(fèi)專項(xiàng)資金(3102016zy019)資助
【分類號】：TN919.81
本文編號：2412487
[Abstract]:A fast coding unit (coding unit,CU) partition method is proposed to solve the problem of high complexity of high efficiency video coding (high efficiency video coding,HEVC) standard coding. Firstly, based on Lagrangian rate-distortion optimization theory and related experimental data analysis, the distortion and bit-rate corresponding to the adjacent CU depth are obtained to satisfy the linear relationship, and the prediction model of rate-distortion cost is established by using this relationship. By using this model, the rate-distortion cost of the next CU depth can be predicted quickly after encoding the current CU depth, and finally, the cost comparison can be used to determine whether it is necessary to continue the CU partitioning. The experimental results show that compared with the HEVC test model HM12.0, the BD-rate of the proposed method increases by 0.2% and 0.6% respectively for low delay and random access coding structures, and at the same time, The coding time was reduced by 33.2% and 38.9%, respectively.
【作者單位】： 西北工業(yè)大學(xué)電子信息學(xué)院;
【基金】：國家自然科學(xué)基金(61371089) 中央高校基本科研業(yè)務(wù)費(fèi)專項(xiàng)資金(3102016zy019)資助
【分類號】：TN919.81
【相似文獻(xiàn)】
相關(guān)期刊論文 前1條
1 趙金濤,吳冬青,廖永畔,莫育俊;SERS天線共振子模型理論對Ag,Cu和Au增強(qiáng)因子的研究[J];寧夏大學(xué)學(xué)報(bào)(自然科學(xué)版);1998年02期
相關(guān)碩士學(xué)位論文 前1條
1 羅霽;HEVC快速CU大小選擇與SKIP模式早期判斷算法[D];西安電子科技大學(xué);2014年
，本文編號：2412487
本文鏈接：http://sikaile.net/kejilunwen/xinxigongchenglunwen/2412487.html
最近更新
教材專著
