Video frequency fire hazard aerosol fog image recognition method based on light stream method

Abstract

The invention discloses a method for recognizing video fire smoke image based on optical flow, which is characterized in that: optical flow of corners and points in smoke motion regions is extracted by a computer based on the video image obtained by a surveillance camera; whether the smoke is fire smoke is judged based on the set threshold by calculating the average and variance of the array consisting all corner and point optical flow speed in the foreground image and the average and variance of the array consisting optical flow speed directions; if a fire smoke is identified, the computer sends orders which controls the alarm to send alarm signals and controls the interlocked fire extinguisher to start the fire extinguishing function quickly; if a fire smoke is not identified, the first step is returned to. The method provided by the invention can not only accurately reflect the motion characteristics of smoke, reducing light interference and impact of similar pixels on the results, but also reduce computation and improve practicality. The method can greatly reduce false alarm rate and quickly and accurately realize fire smoke detection function.

Description

technical field [0001] The invention belongs to the technical field of fire detection, in particular to computer image processing and video fire detection, in particular to a video fire smoke image recognition method based on an optical flow method. Background technique [0002] According to different detection objects and recognition algorithms, visual fire detection technology can be divided into flame image detection technology and smoke image detection technology. Among them, the smoke image recognition detection technology can detect fires earlier and realize early fire detection. A smoke recognition method based on wavelet transform introduced by Pattern Recognition Letters (Pattern Recognition Letters, 2006, 27: 49-58) in the Netherlands, according to the characteristics of blurred edges of the background image under smoke occlusion, using 2 layers of wavelet transform The high-frequency sub-band is used to study the fuzzy occlusion characteristics of smoke. The wave...

AI Technical Summary

Problems solved by technology

A smoke recognition method based on wavelet transform introduced by Pattern Recognition Letters (Pattern Recognition Letters, 2006, 27: 49-58) in the Netherlands, according to the characteristics of blurred edges of the background image under smoke occlusion, using 2 layers of wavelet transform The high-frequency sub-band is used to study the fuzzy occlusion characteristics of smoke. The wavelet transform can better extract the point singular features in the image, but it cannot extract the line singular features well, and there are certain limitations.

At the Suppression and Detection Research and Application Conference (Suppression and Detection Research and Application Conference, SUPDET 2007, Orlando, Florida, March 5-8, 2007) held in Orlando, USA in 2007, Xiong Ziyou et al. proposed a method of extracting smoke using a Gaussian mixture model. motion area, and then identify the method of smoke by analyzing the irregularity of the outline, but because the smoke edge is not obvious and the influence of noise, this method is not effective in the application of complex monitoring conditions

A smoke inspection method based on pixel level introduced by Chinese patent announcement number CN1325903C can identify smoke by detecting the interval of pixel brightness value of the monitoring image. Although the recognition method based on pixel brightness has a faster recognition speed, due to many environmental objects and Smoke has similar color statistics, so false positives are quite high

Transactions of the Institute of Electrical Engineers of Japan (Transactions of the Institute of Electrical Engineers of Japan, Part Dvol.124-d, no.4414-20, 2004) and the Conference on Automatic Target Recognition held in Orlando, USA in 2003 XII, APR 22-24, 2003, ORLANDO, FLORIDA) published two articles about the application of optical flow method to forest fire detection research. In the previous article, it is necessary to find the flame area, which cannot be used for only smoke and the flame is blocked. In the latter article, optical flow is only used to calculate the optical flow of the smoke image, and no feature model is established to distinguish the smoke from the interference source; the optical flow method used in the above two articles has a large amount of calculation and cannot meet the requirements of the early Fire Detection Requirements

Images