{"title":"System anlaysis of power-aware H.264/AVC encoder based on fast inter prediction","authors":"Yeong-Kang Lai, Lien-Fei Chen, Shin-Ping Yang","doi":"10.1109/SOCDC.2008.4815616","DOIUrl":null,"url":null,"abstract":"In H.264 encoder system, the power dissipation of inter prediction occupies a quite large part. For low-power and power-aware system design, efficient fast inter prediction algorithms are the significant techniques to reduce the power consumption. In this paper, a system-level power-aware algorithm based on the early termination scheme of H.264 inter prediction is proposed for the power-aware H.264 encoder. The proposed early termination scheme for H.264 motion estimation is based on the statistical analysis of the transform coefficients. Due to the theoretical model, we develop a power-aware adaptive scheme with multiple thresholds derived from the statistical model to early terminate the motion estimation (ME) operations. Moreover, the proposed inter mode pre-decision is also utilized to early determine the best inter mode at integer motion estimation (IME) stage to substantially decrease the computational cycle counts of the fractional motion estimator (FME) and carry off the motion vector predictor (MVP). According to the experimental results, our system level fast inter prediction design methodology not only preserves fine RD performance but also eliminates the unnecessary operations in both IME stage and FME stage to realize the power-aware H.264 encoder system.","PeriodicalId":405078,"journal":{"name":"2008 International SoC Design Conference","volume":"228 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2008-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2008 International SoC Design Conference","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/SOCDC.2008.4815616","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
In H.264 encoder system, the power dissipation of inter prediction occupies a quite large part. For low-power and power-aware system design, efficient fast inter prediction algorithms are the significant techniques to reduce the power consumption. In this paper, a system-level power-aware algorithm based on the early termination scheme of H.264 inter prediction is proposed for the power-aware H.264 encoder. The proposed early termination scheme for H.264 motion estimation is based on the statistical analysis of the transform coefficients. Due to the theoretical model, we develop a power-aware adaptive scheme with multiple thresholds derived from the statistical model to early terminate the motion estimation (ME) operations. Moreover, the proposed inter mode pre-decision is also utilized to early determine the best inter mode at integer motion estimation (IME) stage to substantially decrease the computational cycle counts of the fractional motion estimator (FME) and carry off the motion vector predictor (MVP). According to the experimental results, our system level fast inter prediction design methodology not only preserves fine RD performance but also eliminates the unnecessary operations in both IME stage and FME stage to realize the power-aware H.264 encoder system.