Fire and Smoke Detection Using Wavelet Analysis and Disorder Characteristics
The fire and smoke monitoring systems are useful in different industry such as military, social security and economical. The recent methods for fire and smoke detection are used only motion and color characteristics thus many wrong alarms are happening and this is decrease the performance of the systems. This research presents a new method for fire and smoke detection through image processing. In this algorithm all objects in an image is considered and then check them to figure out which objects are smoke and fire. The color, motion and disorder are useful characteristics in fire and smoke detection algorithm. Smoke of fire will blur the whole or part of the images. Thus by processing of the video frames, different objects will detect. Due to evaluate the features of objects, the goal objects (fire and smoke) can be defined easily. Twodimensional wavelet analysis is used in the presented method. The results of this research present the proposed features that can reduce the wrong alarms and increase the system performances.
Fire and smoke detection Wavelet analysis Disorder features Color and motion features
AH Rafiee Reza Tavakoli Reza Dianat Sara Abbaspour
Electrical Department Islamic Azad University-Kazeroun Branch Kazeroun,Iran Islamic Azad University-bushehr Branch bushehr,Iran Persian Gulf University Islamic Azad University-Kazeroun Branch Kazeroun,Iran
国际会议
上海
英文
262-265
2011-03-11（万方平台首次上网日期，不代表论文的发表时间）