The Monitoring And Alarm System Of Wire Television Network

With the development of information technology, wire television uses network technology widely, it also has the problem of network security. Because of some illegal invasion of wire television network, interrupting TV programs, criminals spread a large number of illegal information. It causes the state attaches great importance to the safety of the wire television network.This paper’s main research content is the monitoring and alarm system of wire television network, it uses computer digital image recognition technology to improve the traditional monitoring and alarm of wire television network, in order to reduce the manual workload, reduce the burden of network and server.This is done in CTCMC. This company builds a large number of acquisition stations, it makes this development has many samples to test. After the analysis of a large number of illegal screenshots, we determine the basic characteristics of illegal information broadcast—by modifying the EPG electronic programming. We convert EPG ads to logo, and write it into the set-top box hardware. Once the illegal information appears, the logo will appear. Then we spread the equipment to acquisition stations, it makes the illegal information can be monitored immediately and be transmitted to the monitoring center.In this development, we use computer digital image recognition, windows service, Oracle database, multithreading technology and so on. Because there are more than 800 acquisition stations in our country, they conduct real-time monitoring all day long, the number of return images is very large, we need an efficient image recognition algorithm to complete each recognition quickly, to ensure the normal operation of monitoring. So this paper focuses on how to use the computer image recognition to deal with screenshots of low clarity returned by acquisition stations, recognize illegal information in the screenshots. And we use multithreading technology, it improves error-tolerant rate of the system. When a recognition takes a lot of time, it will be suspended by the system, and it will be continued the next idle time. It ensures the normal operation of the system.In order to improve the error-tolerant rate of the system, we combine computer image recognition with artificial recognition, examine the screenshots by multi-level. It makes the screenshots uploaded to the SARFT almost have no error.After test, this system preliminarily meet the SARFT’s requirement of monitoring and alarm of wire television network, initially has the on-line condition. With the on-line of the system, it will save more than 99% of the company’s manual workload, and bring great economic benefits.