Research On A Smoke Concentration Detection Technology Based On Image Processing

This Paper carried out the smoke concentration detection techniques based on imageprocessing method, the formation mechanism and detection methods for artillery muzzlesmoke concentration, and designed an artillery muzzle smoke concentration detection system,for the formation mechanism and detection methods of artillery muzzle smoke concentration.The smoke detection technology is divided into two kinds of image detection andlaser-based detection, respectively, through the study of smoke image and the laserpenetrating to detect the smoke, which is widely used in fire detection, industrialenvironmental protection, the military field. In modern warfare, smoke and flame ofconventional weapons system during launch are produced as little as possible. Therefore,under the premise of ensuring sufficient energy for the artillery, smoke concentrationdetection can provide important reference for the study of propellant.This paper studied the theory of the image smoke detection and laser-based smokedetection, established dynamic model smoke based on Mie scattering theory model and theLambert-Beer law , analyzed the smoke detection technologies based on smoke characteristicsand the multi spectral image, the optical polarization characteristic, and after lightscattering ,carried out the representation and calculation of extinction coefficient, etc. whichproposed the smoke concentration detection program combined image processing and lasercombining with propellant combustion characteristics and the produce mechanism of smoke,designed the smoke detection system based on high-speed photography images of the CCDcamera and lasers, set up the signal acquisition and conditioning circuits, and synchronoustrigger circuit.The experiments selected four kinds of laser of 650nm, 780nm, 850nm, 980nm to build alaser penetrating experimental platform to study the smoke penetration experiment for thefour different wavelength lasers at different optical path , and found that 980nm laser penetrating is the strongest, which provide the basis for the choice of the laser for the muzzlesmoke detection; built the smoke image and laser experimental platform by using the CCDcamera to study the approach of image gray value based on the MATLAB , established therelationship of the gray value and smoke concentration by processing the gray value andanalyzing the transmission of laser and ultimately achieved the purpose of detecting thesmoke concentration.