Design Of Intrusion Detection System Based On Cloud Platform

Visual monitoring and passive infrared signal detection based on image processing are commonly used in intrusion detection.The simple infrared signal detection system has some shortcomings,such as high false alarm rate and unable to obtain evidence.However,the only video surveillance system is not adaptable to the moving objects in the occlusion and dead corner of image monitoring.In addition,the monitoring data generated by multi-point has some problems in the process of transmission and storage,such as excessive network bandwidth and storage space,large amount of data processing and so on.In response to the above problems,this paper develops an intrusion detection system based on cloud platform,which combines image visual monitoring and passive infrared detection.Therefore,an image visual sensor network and an infrared detector early warning network are constructed.In this paper,the cloud server is used to realize the upload and processing of intrusion image and infrared detector monitoring data,as well as the storage and distribution of alarm information And based on the Android platform developed mobile terminal monitoring and alarm software.The following work has been done in this paper:1.Infrared detector warning network : The infrared detector module is developed,and the ZigBee infrared detector network is constructed to realize the detection of human body targets.The developed alarm host will initially filter the infrared alarm data collected by multiple nodes and upload it to the cloud server.2.Image vision sensor network: In this paper,based on Android platform,a monitoring client software is developed to realize remote alarm monitoring.The software receives the infrared and visual alarm information sent by the cloud server,and integrates the infrared and image visual alarm information to form the final alarm result.3.Mobile monitoring client: This paper develops a monitoring client software based on the Android platform and implements remote alarm monitoring.The software receives infrared and visual alarm information issued by the cloud server,and integrates infrared and image visual alarm information to form the final alarm result.The test results show that the intrusion detection system based on infrared detector network and image vision has a low false alarm rate.It has better adaptability to problems such as occlusion and blind area of monitoring,and takes up less network bandwidth.Moreover,it has greater advantages in cost compared with pure video monitoring system.The real-time performance of the system basically meets the application requirements,but the detailed functions of the system still need to be further improved.The subsequent plan is to improve the system from the aspects of infrared detector network space topology and fusion algorithm,so as to further improve the efficiency and practicability of detection.