Research On Infrared Optical Image Enhancement Method Of Leakage Gas

Harmful gases are increasingly widely used in industry and life.These gases may leak into the environment during production,use,storage and transportation,causing great harm to public safety.Harmful gases can be imaged by uncooled focal plane detectors due to their specific infrared absorption spectra.However,gas infrared images have the disadvantages of low signal-to-noise ratio and inconspicuous gas area.Therefore,the infrared image enhancement algorithm of leakage gas and the moving target detection algorithm of gas area are mainly studied in order to make the leakage gas more obvious.Aiming at the problem that the use of infrared image enhancement algorithms results in noise amplification or poor gas leakage enhancement effects,the gas infrared image enhancement algorithm based on guided filters is proposed.First,based on the gas movement between the two frames,the difference between the two frames is used to amplify the gas.The guided filter is used to get a detail layer containing gas texture and noise,and a basic layer containing low-frequency gas information.A gain mask based on the guided filter core is designed to suppress noise while not filtering out the gas texture information in the detail layer.An adaptive histogram enhancement algorithm is used to enhance gas low-frequency information in the base layer.The minimum value through locally weighted scatterplot smoothing is used to divide the sub-histogram,and a sub-histogram range remapping algorithm is designed by local entropy.Finally,the processed detail layer and base layer are used to reconstruct gas-enhanced image.Aiming at the problem of incomplete or incorrect detection of gas area when using the moving target detection algorithm to detect leakage gas,the gas area detection algorithm based on improved PBAS is proposed.The PBAS moving target detection algorithm adaptively controls the update rate and the decision threshold for judging foreground and background,which can suppress the non-gas area foreground caused by noise and camera shake.In order to fully detect the gas area,an adaptive inter-frame difference method controlled by the PBAS decision threshold is designed.At the same time,in view of the problem that a part of long-existing foreground pixels are gradually added to the background model,a update rate controlled by the number of times that the same pixel in consecutive frames is judged as foreground is designed.The detection results obtained by the inter-frame difference and the PBAS algorithm are combined,and then the gas area position is got through the connected domain processing.Finally,the leaked gas is colorized to obtain a gas colorized image.The experimental results show that the gas infrared image enhancement algorithm can enhance the contrast between the gas and the background while reducing noise,and at the same time,the objective indicators such as standard deviation and roughness are improved.The gas area detection algorithm can detect leakage gas to the greatest extent and has a certain robustness to the environment.The colorization algorithm further highlights the leakage gas.