首页> 外国专利> CABAC BINARY ARITHMETIC ENCODER FOR HIGH SPEED PROCESSING OF UHD IMGE

CABAC BINARY ARITHMETIC ENCODER FOR HIGH SPEED PROCESSING OF UHD IMGE

机译:用于UHD IMGE高速处理的CABAC二进制算术编码器

摘要

The present invention relates to a CABAC binary arithmetic encoder for high-speed processing of UHD images, and when the context model value (ValMPS) input from the context modeler is equal to the bin value (binVal), the symbol (MPS) having the highest probability of exiting the current encoding bin (MPS) , Least Probable Symbol (LPS) that determines the least likely probability of exiting the current coding bin if ValMPS is not equal to the bin value, and, if the ValMPS is not equal to the bin value, the current coding bin When the range of the range (Range) is less than a certain range, the range update stage to perform the normalization, and outputs the range of the normalization (Cnt_RenormE) and the MPS range (rMPS) required for calculating the minimum boundary value (Low) of the range. Add the value of rMPS and the Low to obtain the ivlLow value for the next call, and it is necessary to generate a bitstream for the currently encoded bin through the ivlLow value and the Cnt_RenormE. By yirueojim the bit stream through the information bits for the blank to a minimum threshold update stage and the current encoding and outputting the information bits to the bit stream generation stage for outputting, it is possible to reduce the maximum delay time (Critical Path) optimally.
机译:本发明涉及用于UHD图像的高速处理的CABAC二进制算术编码器,并且当从上下文建模器输入的上下文模型值(ValMPS)等于bin值(binVal)时,符号(MPS)具有退出当前编码bin(MPS)的最高概率;如果ValMPS不等于bin值,并且如果ValMPS不等于bin,则最小概率符号(LPS)确定退出当前编码bin的可能性最小。 bin值,当前编码bin当范围(Range)的范围小于某个范围时,范围更新阶段执行标准化,并输出所需的标准化范围(Cnt_RenormE)和MPS范围(rMPS)用于计算范围的最小边界值(低)。将rMPS的值和Low相加以获得下一个调用的ivlLow值,并且有必要通过ivlLow值和Cnt_RenormE为当前编码的bin生成位流。通过将空白信息位的比特流传输到最小阈值更新阶段,并将当前信息编码并输出到比特流生成阶段以进行输出,可以最佳地减少最大延迟时间(关键路径) 。

著录项

  • 公开/公告号KR1021097680000B1

    专利类型

  • 公开/公告日2020-05-13

    原文格式PDF

  • 申请/专利权人

    申请/专利号KR1020170127846

  • 发明设计人 류광기;조현구;

    申请日2017-09-29

  • 分类号

  • 国家 KR

  • 入库时间 2022-08-21 10:58:02

相似文献

  • 专利
  • 外文文献
获取专利

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号