基于多维矢量矩阵的运动自适应扫描视频压缩编码研究

The study of the adaptive scanning video compression coding based on multidimensional vector matrix

针对现代视频图像数据量较为庞大,导致传输带宽及存储空间增加,以此无法实现和视频图像数据量提高趋势成正比,所以就要实现视频数据的压缩。保证视频不失真,从而得到最小的单位形式比特数,以此提出基于多维矢量矩阵的运动自适应扫描视频压缩编码,其创新了传统多维变换运算的方式,并且融入了多维矢量矩阵正交变换的压缩技术,表示此理论具有一定的优势。首先,对多维矢量矩阵理论进行了分析,之后分析了时间轴中系数分布的规律,表示时间轴中的视频运动信息会影响到视频多维适量矩阵正交变换系数能量分布,根据系数和平面之间的距离设计全新的扫描算法,也就是运动自适应扫描算法。最后对其进行实验,通过实验表示,基于多维适量矩阵的运动自适应扫描视频压缩编码比传统自适应扫描具有进步性。

In the modern video image data quantity is relatively large,lead to increase the transmission bandwidth and storage space,this cannot be achieved and the video image data quantity is directly proportional to the increase trend,so to realize the compression of video data. Ensure the video is not distortion,resulting in the smallest unit in the form of bits,thus put forward based on multidimensional vector matrix scanning motion adaptive video compression coding,the innovation of the traditional multidimensional transform operation way,and entered into the compression technology of multidimensional vector matrix orthogonal transformation,says this theory has a certain advantage. First of all,analyses the multidimensional vector matrix theory,then analyzes the coefficient distribution regularity in time axis,said the timeline of the video motion information will affect the multidimensional moderate video matrix orthogonal transformation coefficient of energy distribution,based on the distance between the coefficient and plane design new scanning algorithm,namely motion adaptive scanning algorithm.Finally,the experiment was carried out to show that the adaptive scanning of video compression coding based on the multi-dimensional appropriate matrix was more progressive than the traditional adaptive scanning.