Research On Indoor Intelligent Firedetection Based On Video

With the advancement of science and technology,video monitoring technology has been widely used.Compared with traditional monitoring technology,video monitoring technology has many advantages such as being intuitive,real-time and reliable.In particular,by inquiring into historical data and finding out the cause of an alarm event and defining the responsibility of an accident,it plays an irreplaceable and active role in the prevention of later events.This paper mainly studies the indoor flame detection technology based on video.This paper studies the design of flame detection algorithm and the concrete realization process on the basis of fire incident in a university laboratory.Get the background image first and because of the slow variation of the monitoring scene,the background image is obtained by means of the multi-frame image averaging method.At the same time,the image is based on the custom block to improve the reliability of the background.The difference between the real-time image and the template data is compared with the appropriate threshold to update the background.Secondly is the real-time extraction of the moving target.Based on the background image,the current frame and the background frame difference method are used to detect the moving target.In order to eliminate some interference areas and improve the efficiency of the post processing,the extracted motion targets were treated with binary block,connected domain mark,grayscale statistical consistency processing,motion accumulative processing,morphological processing and other image pretreatment methods.Subsequently,the segmentation targets were further screened and confirmed from the circular characteristics,area growth characteristics and frequency band characteristics of the flame,thus improving the reliability of the algorithm.Finally,each characteristic of the flame is extracted as a weak classifier,and the Adaboost algorithm is used to fuse the weak classifier into a strong classifier to realize the rapid detection of fire.This paper develops a flame detection system based on VC6.0 platform.Through experimental verification,the algorithms used in this paper can accurately and quickly identify the flame generated by fire.The feasibility and practicability are high,and it is of great significance to the laboratory fire warning.