---

3.6 Video Analysis

版权声明: 转载请注明出处 http://www.codingsoho.com/

视频监控和分析已经广泛应用在各个领域，如车牌识别、车流量统计等。为了避免要求视频摄像头进行本地视频分析导致的配置增强、成本增加问题，目前通常将监控视频数据上传到远端云计算中心的视频服务器，由视频服务器进行相关的分析处理及最终存储。然而，大量未处理的原始视频的直接上传对回传带宽以及视频服务器存储空间提出了很高的要求，特别是视频中哪些无用信息，直接上传对传输资源是极大的浪费。

基于MEC的计算和存储能力，可以给出基于MEC的视频分析框架。由部署在MEC服务器上的视频管理和分析应用负责完成视频的分析处理，如视频压缩、事件监测、视频剪辑等。然后，视频管理及分析应用根据实现配置上报监测的特定事件、压缩后的视频或者剪辑好的视频片段。此时，处理后的视频可以有效降低对回传带宽的消耗。

Video monitor and analysis is widely used in all the industries, such as LPR - vehicle license plate recognition, statistics of the vehicle flow, ect. Considering the cost to high configuration requirement of local video analysis, the common solution is to upload the video to remote cloud video server for relevant analysis and final store. However, load of untreated raw video upload has high demand on the backhaul bandwidth and video server storage space, especially for the se useless information, it’s enormous waste of transmit resource to upload these information.

We can provide video analysis framework based on the compute and storage capability of MEC, . application deployed on MEC server will be responsible for video management and analysis, such as video compression, event monitor, and video edit. Then the applications report the specific event, compressed or edited video according to predefined configuration. The solution can effectively reduce the backhaul bandwidth cost.