摘要
文中提出了一种基于Fourier-Mellin变换的视频压缩算法。该算法应用Fourier-Mellin变换,获取Fourier-Mellin变换系数并矢量化,得到了图像的几何特征不变量,提取视频中相邻图像帧之间的平移、旋转和缩放三组运动参数。在视频传输的过程中仅需传送参考帧和运动参数数据包,就能较好地恢复出原视频文件。实验结果表明,该算法运算量小、压缩比高,恢复原视频文件效果较好,能有效地改善传统视频压缩文件中冗余信息较多而引起传输速率慢的缺点。
The compression algorithm of surveillance video files was proposed based on Fourier-Mellin transform. In the paper,the coefficient of Fourier-Mellin was achieved to vectorization with the new algorithm,the surveillance video contains contiguous image frames,three groups of movement parameters were extracted before transmission depend on the constant geometric dimension,including the translation,rotation and scaling, it can more better to restore the original. This compression algorithm after supervising and controling video,parameter packet and reference were only transmitted in process. The experiment result expresses that the algorithm is viable,the small efficient calculation and the high compression ratio. therefore,it was being useful to overcome a question of the transmission redundancy information in traditional compression video leading to the characteristics of speed slowly.
出处
《信息技术》
2016年第3期73-75,79,共4页
Information Technology
基金
商洛学院科研基金资助项目(14sky001)
