Fire smog detection method based on motion characteristics and convolutional neural network

Abstract

The invention relates to a fire smog detection method based on motion characteristics and the convolutional neural network. Through reading a video file, a first image is stored as an original image,and smog detection on each frame of the video is carried out; firstly, the original image is added to background update as reference, a background model is further established, secondly, a foregroundimage is extracted through a difference method, the foreground image is filtered through a dark channel threshold image to acquire candidate smog areas, lastly, a depth convolutional neural network model after training is loaded to automatically extract high-level characteristics of the candidate smog areas, and whether the candidate smog areas are smog areas is determined according to extracted characteristic vectors. The method is advantaged in that the channel prior knowledge is added to motion foreground detection, common interference is effectively filtered, environment adaptability of adetection method is improved, the convolutional neural network is used for carrying out characteristic extraction of smog images, and detection efficiency is substantially improved.

Description

technical field [0001] The invention belongs to the technical field of fire monitoring, and in particular relates to a fire smoke detection method based on motion characteristics and a convolutional neural network. Background technique [0002] Tens of thousands of fires occur every day all over the world, causing hundreds of casualties and destroying large areas of forest vegetation. Fire seriously threatens the safety of human life and property and the natural ecological environment. Fires are often sudden, wide-ranging, and difficult to deal with. Therefore, real-time monitoring and timely warning of fire has become particularly important. Early detection of fire is the key to minimizing damage, because once the fire spreads, it will be difficult to control. Generally, the flame is small at the initial stage of a fire, but the smoke is obvious, so the detection of fire smoke is an important basis for judging whether a fire has occurred in time. [0003] Traditional fi...

AI Technical Summary

Problems solved by technology

Toreyin et al. (Pattern Recognition Letters, 2006, 27:49-58) conduct smoke detection based on wavelet analysis. The principle of this method is that when smoke is generated, the edge of the background will be blurred, and the energy of the high frequency part will be reduced. The intensity component attenuates and the brightness value decreases. Due to the need to analyze the background contour of the scene, the scope of application of the algorithm is limited.

Piccinini et al. (15th IEEE International Conference on, ICIP 2008, California, October12-15, 2008) segmented the smoke area by online modeling of the ratio of foreground energy and background energy. Better, but the calculation complexity is high, and it is generally difficult to guarantee real-time performance

Fujiwara et al. (International Symposium on Communication and Information Technologies, SUPDET 2007, Orlando, Florida, March5-8, 2007) extract the smoke area from the image according to the self-similar fractal theory. For low-contrast, blurred smoke images, the extracted Fractal features are not stable enough

Chinese Patent Publication No. CN101441771A introduces a video fire smoke detection method based on color saturation and motion mode. The above characteristics are difficult to obtain continuous statistics in the scene of distance, thus limiting the scope of its use

Images