Font Size: a A A

Research And Design On The Security System And The Video Behavior Analysis Method

Posted on:2015-08-06Degree:MasterType:Thesis
Country:ChinaCandidate:S N ChengFull Text:PDF
GTID:2298330467463872Subject:Electronic Science and Technology
Abstract/Summary:PDF Full Text Request
Security system attracted more and more attention at home and abroad with the economic and cultural development. Traditional security systems software is not compatible with different kinds of manufacturers, and it is not maintainable for developers because of its C/S structure; meanwhile, the security attendants usually monitoring the video wall with7*24hours and searching for abnormal scenes from plenty of videos.This paper firstly analyzes the trend of the security system and the actual needs of a project which this paper based on. In order to solve the sensor compatibility issues and the interaction protocol conversion and parse issues, this paper divides the system into three layers. We define the format of interactive communication protocol and make use of the middleware technology. To make the infrared information, including the temperature and humidity information, video information, work together to provide a more comprehensive alarm information, this paper developed a range of information process flow.Secondly, in order to take full advantage of video information, not limited to the real-time viewing and recording, the crowd density detection algorithm is studied in the paper. We ultimately select a pixel-based statistical and texture analysis based crowd density detection algorithm after analyzed and compared. Then, to make the crowd density detection algorithm useful and effective in an actual security system, real-time processing of video sequences technology is discussed in the paper. We select the key-frame extraction technology, buffer and sliding window technology as processing method for the real-time video sequences. Then, the processed video sequence will be sent into the crowd density detection module, the results is used for the crowd density warning.Finally, in order to verify the above scheme can solve the compatibility of the sensor, achieve the purpose of collaborative alarm, and real-time detection of crowd density, we deployed the system designed in this paper. Through a few practice, we verify the feasibility of the security system. In the end, the paper analyzes and summarizes the shortcomings and proposes further research directions.
Keywords/Search Tags:security system, crowd density detection, real-timevideo sequence processing
PDF Full Text Request
Related items