The Research On The Key Technology Of Real-time Monitoring Fire Based On Internet

The modern buildings should possess the function of automatic fire detection, automatic alarm, automatic control. In order to avoid or reduce the losses of property when fire happened, we must forecast early fire. Compared with traditional forecast method in fire forecast, digital image processing method can improve the precision of predicting, shorten predicted time, offer more abundant fire information etc. With the rapid development of computer technology, network technology, communication technology, we can real-time monitor urban fire at a low cost in large range.At present, there is an undefined delay in Internet, which is disadvantageous to real-time performance of the fire alarm system. This paper further investigates TCP/IP protocol, IP routing principle and router operational principle. This paper analyses the network delay in detail. Different with the way of normal router in transmitting all data packages fairly, this paper put forward a new dispatch algorithm, which can identify different data packages of demanding different real-time performance and transmit the data packages of demanding high real-time performance by jumping queue intelligently. That will decrease warning time delay and improve the dependability of important data transmitted by router. This paper offers a new kind of method to satisfy a high real-time system at a low cost.As a new branch of mathematics and with a great ability to describe and study natural phenomenon, fractal geometry has attracted a good deal of attention and has been applied in more and more fields. As an important characteristic for fractal geometry, fractal dimension describes the roughness of natural surfaces. This paper introduces fractal brown model and fractal dimension into the texture analysis of flame images, which can forecast early fire and reduce the losses of property when fire happened.As the application of the network being more and more extensive, the network security becomes more and more important. Infonnation encrypts can prevent information from being eavesdropped while transmitting in the open network. According to present application, the traditional cryptographic method is easy to crack. At present, some non-traditional methods is studied in the information encrypt. The chaos theory is one ofthem that is adopted and acquired the extensive application. Having sensitivity dependent on initial conditions and pseud-randomness, we can use chaotic sequence as stream cipher. This paper offers a compound chaotic system based mutiple model used as sequence cipher, which has a high encryption speed, a large key space and is difficult to crack.