Design And Implementation Of Army Remote Video Surveillance System Based On Internet Of Things

With the continuous updating and iteration of network technology,monitoring technology has also undergone earth-shaking changes.Because the monitoring system has good practicability and applicability,the monitoring system is getting more and more attention and recognition.Remote monitoring in the monitoring system has the technical characteristics of unattended and all-weather on-the-job and has been widely used in civil,military and other fields in recent years.With the rapid development of network technology,especially the Internet of Things technology,it is urgent to develop a military remote video monitoring system based on the Internet of Things for the army to establish an integrated system of the Internet of Things as soon as possible.With the continuous update and iteration of network technology,the monitoring technology has undergone earth-shaking changes.Due to the good practicability and applicability of the monitoring system,the monitoring system is increasingly valued and recognized by everyone.Remote monitoring in the monitoring system has the technical characteristics of unattended,all-weather on-duty,and has been widely used in civil and military fields in recent years.In the context of the rapid development of network technology,especially the Internet of things technology,it is extremely urgent to develop the remote video monitoring system based on the Internet of things for the army to establish the integration system of the Internet of things as soon as possible.At present,the pedestrian flow statistics based on computer vision is the mainstream technical scheme in this field,but how to realize the pedestrian flow statistics system with high accuracy and good real-time performance is still a difficult problem in practical research.In this paper,dirichlet prior function and maximum likelihood estimation are introduced to reduce the weight.The new improved model can not only eliminate the gaussian model that does not meet the classification conditions,but also prevent the gaussian model with negative weight,and can realize the adaptive selection of the classification number of the gaussian model.In the common monitoring system,not only lacks the effective intelligent monitoring function,but also has the operation inconv enience,the management inconvenience difficult problem.On the one hand,this paper improves the existing gaussian human behavior recognition model and proposes a new human behavior recognition algorithm to preprocess relevant monitoring images in video m onitoring,so as to achieve the purpose of intelligent management.On the other hand,aiming at the above problems in common monitoring systems,combined with the theoretical basis of remote monitoring and the Internet of things technology,a remote video monitoring system for troops based on the Internet of things is designed.This system can realize front-end file processing of monitoring equipment,obtain continuous video information of monitored objects,and have business functions such as remote operation,equipment restart and shutdown.Based on Exynos processor and Web server Boa,the design and implementation of remote monitoring control system for embedded and ARM9 development board are introduced in detail.The development board is used as the server,running the Linux operating system to loop through the main function and transfer the video data detected by the image acquisition terminal(camera)to the Web server.In this paper,Linux real-time operating system(ubuntu12.04)is transplanted on the ARM9 processor core-based 4412 development board,and the Boa Web server is transplanted.