| As video surveillance system is widely applied in community and security, storage and analytics of video big data is a very useful subject. Storage and analytics of video big data include the way to store and analyze data, and analytics between video data and other format of data.Since analytics of big data should cover both real-time and accuracy, there must be a way to quickly analyze big data. Also, analytics of video data usually focus on human body, car, which has a complex motion, appearance and environment. This makes the analytics/modeling very hard to build. Also, the way to store and analyze data should cover efficiency and cost saving. This article research on storage method of video big data, come up with edge a combination of cloud and central cloud to store and analyze data. Edge cloud is based the lightweight vitalization system, called Docker. The central cloud is based on Hadoop + Mapreduce, which enables the real-time data fetch and analytics. And this acticle provides an introduction on edge cloud, Docker, central cloud, Hadoop, plus Haddop Mapreduce. And compare tradition way with what this article presented, shows pros and cons of these two systems. According to the performanc testing, we can see that it’s more igile and safe by using center-edge cloud system, comparing to the traditional method. Especially for storage and analyze of big data, its efficiency and safty are many times than the traditional way.This article introduces edge cloud that is based on light weight virtualization system Docker, and together with central cloud that is based on HDFS(Hadoop Distributed File System) and MapReduce. Configured storage/analytics model by using edge cloud/central cloud, verified the performance of the system. Also this article introduced the future of video big data analytics. |
PDF Full Text Request