A Method Of Constructing Structured Video Information

With the deepening of Sky-net Project, Peaceful City, the amount of video information is increasing continuously. To find the information we need in massive video data, it’s like fishing for a needle in the ocean and also brings great challenges to the video surveillance. The traditional huge-crowd strategy makes the video surveillance in state of prison without control because of low efficiency and it is easy to miss key goals. How to solve the problem of intelligent storage of video information and efficient retrieving has become very hot and difficult.At present, the processing method of dealing with intelligent video is based on the analysis of image content and it is widely used at home and abroad. The method plays a role in video retrieving, but the efficiency of video retrieving is still not so high and the accuracy also changes with the environment. To solve this problem, we design a method to construct structured video information. Combine unstructured video information with traditional sensor information to make up structured information in order to solve the problem of the utilization efficiency of video information in security surveillance.Firstly, the paper introduces the principle of building structured video information. In two aspects including JPEG image files and MKV video files introduces the method of inserting IOT sensor information into JPEG files or generating description file for video files and introduces the method of retrieving and process for these two kinds of files. And then design a system which is based on structured video information. The system includes intelligent video capturing andencoding device, real-time decoding and playing device and the comprehensive server for managing and retrieving.Secondly, the paper has developed a smart video capturing and encoding device prototype and a real-time decoding and playing device prototype. The paper also described the design and development process of these devices in detail including the introduction of device, the partition of module, the process of program and the color space conversion involving in encoding and decoding.Thirdly, the paper has developed software to manage devices and retrieve information which is on the comprehensive server. The paper also described the software’s module division, work flow and GUI interaction design. Through the software we can achieve the goals which are intelligent storage and quick retrieving based on structured video information.The paper aims to provide a method to construct structured video information, in order to improve the utilization efficiency of video information in the field of security surveillance. It provides information assurance for the field of public safety and put forward a new solution for the further effectively utilize of video information which is captured by other industry at the same time. The paper fully considers the expansibility of the system when designing this system. It can be used in security protection, industrial control, environmental monitoring, medical monitoring, and many other fields.