{"title":"A high performance motion vector processor IP design for H.264/AVC","authors":"K. Yoo, S. Park, Hyunsuk Ko, K. Sohn","doi":"10.1109/ISCE.2008.4559504","DOIUrl":null,"url":null,"abstract":"In this paper, the world's first hardware design of the motion vector processor of H.264/AVC and its FPGA implementation are presented. It aims at a low-cost high-throughput design for HD1080 (1920 times 1088) at 60 frames per second (fps) in High Profile (HP) H.264/AVC codec with Level 4.2. For this, deterministic processing loops control scheme and a novel 4 times 4 processing order substituting for the conventional double-Z one are presented to attain a high-throughput design. In addition, for maximizing hardware utilization and getting a low-cost design, two processing elements dedicated to motion vector derivation are presented. The proposed design was realized with 41 K logic gates and 4,608 bits SRAM at 266 MHz and was completely conformed for Allegro compliance bitstreams on an FPGA platform.","PeriodicalId":378486,"journal":{"name":"2008 IEEE International Symposium on Consumer Electronics","volume":"1 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2008-04-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2008 IEEE International Symposium on Consumer Electronics","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ISCE.2008.4559504","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 this paper, the world's first hardware design of the motion vector processor of H.264/AVC and its FPGA implementation are presented. It aims at a low-cost high-throughput design for HD1080 (1920 times 1088) at 60 frames per second (fps) in High Profile (HP) H.264/AVC codec with Level 4.2. For this, deterministic processing loops control scheme and a novel 4 times 4 processing order substituting for the conventional double-Z one are presented to attain a high-throughput design. In addition, for maximizing hardware utilization and getting a low-cost design, two processing elements dedicated to motion vector derivation are presented. The proposed design was realized with 41 K logic gates and 4,608 bits SRAM at 266 MHz and was completely conformed for Allegro compliance bitstreams on an FPGA platform.