Application Of Information Processing Technology For Cloud Video Monitoring In Food Traceability

With the continuous development of the society and improvement of people’s livingstandard, as well as constantly exposed food safety problems, people are getting more andmore conscious of food safety. Due to its function of regulating production by monitoring,deterring and tracing, video monitoring system has been used in the field of food safety. Thecomputing capacityof cloud computing and the storage capacity of distributed file system canbe used to meet the need of building and maintaining mass storage space requested byincreasing storage and query demands, which are caused by the enlargement of area andprolongation of time of the video monitoring system.A platform dealing with cloud video data will be designed through studying the HDFSsystem of the cloud computing system and video monitoring system and taking into accountthe real need of Yinbao Company, in order to achieve efficient, stable and costly-effectivestorage of the increasing mass monitoring video and real time transport of these video data,and at the same time, make them traceable and thus playing its role of monitoring andregulating production.The thesis deeply studies the distributed processing structure, storage means and datareading and writing process of specific fields of Hadoop file system which are especiallyuseful for food safety. Some structural fine tuning was made after referring to manydistributed file systems put in practical use in the process of production, to make the HDFSsystem more applicable in terms of storage and query and more stable for the end of making itmore suitable for video monitoring in the field of food safety. Considering the real need ofsmall and medium-sized companies, structural improvement was firstly made to deploy thesuitabledistributive Hadoop cluster for Yinbao Company owning many plants. Cachingmechanisms were added to meet the need of real time monitoring. An assistant database wasset up on the transaction server to match the video documents and the assistant informationfor the purpose of making it easy to trace data and to access the system.The Namenode wassegmented to effectively reduce Namenode bottlenecks and prevent a systematic failurecaused by any Namenode failure. Based on the above studying, a cloud video information processing system featured by“real time monitoring, video tracing, security and reliability and dynamic scalability” will bedesigned to provide new solution to the video monitoring technique used in the field of foodsafety.