Research On Smoke And Fire Detector Based On Digital Video Processing

In recent years, the number of fire disaster has increased. Safety becomes more and more important. The faster fire is detected, the less we will lose. Therefore, fire detector is in demand. However, almost the traditional fire detectors have poor visibility and can’t perform in real-time. Although the monitoring systems are visible, due to the limitation of monitoring, the existed alarm system’s false rate and miss rate are higher than the traditional fire detectors. In order to solve the above problems, a kind of smoke and fire detect system is given out in this paper. On the premise of high detecting rate, this system uses smoke and fire detection algorithm to detect fire. Experimental results show that the system can guarantee the real-time alarm of fire in most kinds of situations. The main work in this paper can be read below:1) A HSI and movement accumulation characteristics based flame detecting algorithm is proposedFirstly, video should be preprocessed to wipe out noise and adjust in gray levels. Then, the color characteristics and dynamic characteristics of the flame are extracted. Finally, the flame area in the image can be fixed, and the alarm signals are given out. The standard of the fire degree is GB15631-2008. Results show that this algorithm can run in real-time and robust to disturbance.2) A video fuzzy degree and clustering feature based smoke detecting algorithm is proposedFirstly, video should be preprocessed to wipe out noise. Then, extract the fuzzy degree of the image, including contrast, uniformity and energy, the extraction of characteristics of clustering, and ultimately determine the existence of smoke in the image. Experiments show that, the algorithm can roughly fixed the position of the smoke area, and given out early fire warnings.3) A smoke and fire alarm system is designed and implementedThe system is divided into four modules, and they are:image preprocessing, the smoke detecting, the flame detecting arid a man-machine interaction. Image preprocessing module is used to improve the quality of the images. Smoke detecting module and the flame detecting module are used for monitoring the flame and smoke in the image, the man-machine interaction module is used to set parameters, saving log records, and giving out alarms and warnings.Experiments show that, this system can ensure the existence and the location of the flame and smoke in videos. The grade of the fire can be measured out. This system has satisfied the general image fire detection system’s requirements.