Research On Dynamic Linkage Video Surveillance System Based On Wireless Sensor Network

With the development of technology and the society, people's awareness of security has been improved. Video surveillance systems are also put into service in more and more fields. Traditional video surveillance systems have a lot of disadvantages. They are very difficult to be installed and the wiring installation is very complex. Traditional video surveillance systems are usually closed systems, which can hardly take alarm automatically and be cooperated with other systems. Wireless sensor network is a new technology which comes out in recent years. It combines a lot of communicable low power consumption sensor nodes into a big wireless network that can measure and process information in a certain area and the data it measured can be delivered to the final users by the multi-hop and self-organized network.This paper aims at the shortcomings of current video surveillance systems and proposes a new method for intelligent video surveillance systems based on the wireless sensor network technology. With this new method, traditional video surveillance systems become able to cooperate with wireless sensor networks. The applicable scope of video surveillance systems is expanded and the accuracy of the automatic alarm of video surveillance system is improved. The surveillance systems can provide users a lot of useful information automatically, such as the real-time best fire escape route, etc. The safety of the deployment area is improved.According to this method, vibration sensor nodes are deployed on the doors and windows of the building, process and integrate the output data of the three-axis g-sensors within the vibration sensor nodes. The integrated output data of g-sensors pass through the Kalman filter, the low pass filter and the mean filter and the high frequency noise will be eliminated. The MCU recognizes the illegal intrusion according to the characteristic points of the g-sensor output waveform and turns on the relative cameras. An event message will also be sent to the user. This method improves the alarm accuracy of the system and covers the shortcomings of the traditional video surveillance systems whose false alarm rate in low light conditions is very high. This method also uses smoke sensors and flame sensors for real-time monitoring of fire within the building. In case of fire, the fire can be automatically reported to users and relative cameras will be turned on. The system will acknowledge the congestion level according to the measurement of pyroelectric infrared sensors and the real-time images captured by the cameras deployed in every passages in the building. Combined with smoke and fire location reported by the smoke and flame sensors, system will calculate the best real-time escape route and send to the ground safety exit arrow indicator for displaying and showing to the users. This improves the safety of the building.This paper designs the specific hardware circuit and software algorithms and implements the method. The performance of the algorithms is verified by experiments. Experimental results show that this method can accurately detect the illegal intrusion into the building and the occurrence of the fire. The accuracy rate of this method is higher than traditional video image processing methods. The method has a high accuracy rate of congestion detection and indoor smoke detection, and the best fire escape route calculation has an advantage with its features of low latency, real-time and high reliability.