节点文献
基于人头部图像分割的视频信息压缩算法
Compressed Computation of Video Information Based on Human Head Region Image Segment
【摘要】 为提高监测系统中的视频压缩比,在研究监测系统中视频图像特性的基础上,提出了针对人的头部分割的视频压缩算法:首先,用运动信息分割出人体轮廓;其次,运用形态数学的方法实现人头部区域分割;最后,用小波变换和描述人头部对象的差分链码技术来实现高效的视频压缩。本文提出的视频压缩算法比MPEG算法提高了图像的PSNR近3dB。
【Abstract】 Based on the research on video picturing characters in monitoring system, this paper introduces the video compressed computation of human head as the following. First, some movement information is applied to outline a human body; second, the diagram of the human head is segmented with morphological mathematics; finally, a high compression rate is achieved by combining wavelet transformation with different chain codes. The video compressed computation has improved PSNR nearly 3dB compared with MPEG standard at the same data rate.
【关键词】 人头部区域分割;
形态数学;
小波变换;
【Key words】 human head region segment; morphological mathematics; wavelet transform;
【Key words】 human head region segment; morphological mathematics; wavelet transform;
- 【文献出处】 计算机与现代化 ,Computer and Modernization , 编辑部邮箱 ,2005年10期
- 【分类号】TN911.73
- 【下载频次】102