Linux下多线程的视频图像平滑度评价算法

An Evaluation Algorithm of Smoothness in Video Images with multithreading Technology under Linux

视频图像的平滑度评价算法是对视频质量评价的重要指标。平滑度与视频运动信息相关,具有计算量密集和实时性特点。随着流媒体业务的快速发展,传统的评价算法不能满足现状,为了对视频图像质量进行大规模处理,在现有的平滑度评价算法上,提出在Linux环境下处理视频图像,采用多线程技术实现该算法程序。试验结果证明:在Linux环境下实现的多线程算法比传统的串行算法提高了10倍的效率,在大规模计算密集型任务有具大的潜力。

Evaluation algorithm of smoothness in the video images is an important indicator of video quality assessment.Smoothness is related to the video motion activity and has the features of intensive computing and real time.With the rapid development of streaming media service,the traditional evalution algorithms cannot meet the current situation.In order to make the quality of video images processed massively,based on the existing algorithms of smoothness evaluation,this article proposes video images be processed under the Linux environment and have the algorithm realised with the technology of multithreading.The result shows that the speed of multithreading algorithm in the Linux environment is 10-time of the traditional serial algorithm and has great potential in massive compute - intensive tasks.