首页> 外国专利> Global and Local Moving Adaptive Image Coding Device and Method

Global and Local Moving Adaptive Image Coding Device and Method

机译:全局和局部运动自适应图像编码装置及方法

摘要

The present invention relates to a video code, apparatus and method of a video telephone using a motion compensation type video compression technique adapted to local movement (segment unit movement) and global movement (object unit movement).;In order to encode a difference signal between a current video signal and a previous video signal in a frame unit having a plurality of segments, the global and local moving adaptive image coding according to the present invention first encodes the current video signal and the previous video signal, And extracts a motion vector group by estimating motion vectors of all the segments in the frame using a predetermined motion estimation algorithm as an input. Next, a moving region within the frame is extracted by using a predetermined moving region discrimination technique using the current video signal, the previous video signal, and the motion vector group as input, and the extracted number of moving regions is divided by a predetermined threshold value It is determined whether the movement generated in the frame is the global movement, which is the object-based movement, or the global movement, rather than the global movement. And generating the prediction signal by extracting global motion information, which is motion information of an object unit, with the previous image and the motion vector group as inputs, and moving the previous image by the global motion information, If it is determined that the motion is locally shifted, the previous image and the motion vector group are input, and each segment of the previous image is shifted by a motion vector of each segment to generate the prediction signal, and the prediction signal is encoded.
机译:本发明涉及使用运动补偿型视频压缩技术的视频电话的视频代码,设备和方法,该技术适用于局部运动(段单位运动)和全局运动(对象单位运动)。在具有多个段的帧单元中的当前视频信号和先前视频信号之间,根据本发明的全局和局部运动自适应图像编码首先对当前视频信号和先前视频信号进行编码,并提取运动矢量通过使用预定的运动估计算法作为输入来估计帧中所有段的运动矢量来进行分组。接下来,使用预定的运动区域判别技术,使用当前视频信号,先前的视频信号和运动矢量组作为输入,来提取帧内的运动区域,并且将所提取的运动区域的数量除以预定的阈值值确定框架中生成的运动是全局运动(是基于对象的运动)还是全局运动,而不是全局运动。并且,通过以前一图像和运动矢量组为输入,提取作为对象单位的运动信息的全局运动信息,并通过该全局运动信息来移动前一图像,来生成预测信号。将图像局部移位,输入先前图像和运动矢量组,并且将先前图像的每个片段移位每个片段的运动矢量以生成预测信号,并对预测信号进行编码。

著录项

  • 公开/公告号KR940017888A

    专利类型

  • 公开/公告日1994-07-27

    原文格式PDF

  • 申请/专利权人 배순훈;

    申请/专利号KR19920025240

  • 发明设计人 천상훈;

    申请日1992-12-23

  • 分类号H04N7/16;

  • 国家 KR

  • 入库时间 2022-08-22 04:37:34

相似文献

  • 专利
  • 外文文献
  • 中文文献
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