| As we all know，fire disaster is a common serious problem which the wholepeople facing with‘It has made lots of damage to the people society. The seriousdamages proved, fire disaster is a kind of disaster which happens continually andlasts for a long time.To avoid the large accident caused by fire, people used lots of early ifredetection, these methods almost used sensors to detect the humidity, temperatureor smoke in the air to make alarm. The biggest disadvantage of these detectingmethods is the limit of detecting range. So these traditional methods c’ant be usedoutside. They also can’t provide other detail information. The wrong alarmalways happened.In the recent years, many researchers proposed the ifre detecting method basedon video. This one can solve the traditional method’s problems. People analyzethe images’ changes on color, time and space to ifnd the ifre in video. The newmethod has lots of advantages and large applying future.U’sing the fire videos as studying object, with the videos running, colors andtwinkle character, the paper proposed a detection based on video sequences.The paper ifrst found the running range in the video; then make a color modelto analyze the running region. Later, we used the wavelet method to analyze thef’ires flick frequency. To improve the detection accuracy. The paper also addedother methods to analyze the videos. If the video had fire in it, the result wouldshow that the ifre happened and make alarms. |
