Research On The Sniper Localization And The Bullet Parameters Estimation Algorithm Based On The Gun Shooting Sound Model

With the development of modern society, social security and counter-terrorism situation have been paid great attention. The frequently occurring gun shoot events, have brought social unstable factors to many countries. Guns have also become the major threat to private property security. Fortunately the existing gunfire localization systems can quickly determine the orientation of the sniper and the bullet parameters. It has an important application value to maintain a stable social security environment in cities as well as keep the soldiers alive on the battlefields. As acoustic detection based gunfire localization systems have the advantages of strong anti-interference, great feasibility and low cost, they are widely applied in anti-sniper detection system.A conventional firearm uses a confined explosive charge to propel the bullet out of gun barrel. The hot, rapidly expanding gas causes an acoustic muzzle. Then the bullet traveling at supersonic speed in air will cause an acoustic shockwave. Gunfire localization systems will locate sniper’s position and estimate parameters of bullet according to the detection and identification of shock wave and muzzle blast.This thesis finally estimate the bullet’s trajectory, speed, length, diameter and the shooting locations effectively through analyzing signal characteristics of shock wave and muzzle blast which are produced by gunfire. The Difference of Gaussian (DoG) method can detect the shock wave and muzzle blast correctly. We can obtain shock wave’s length, amplitude, slope and the time difference between shock wave and muzzle blast. The signal delay between every two microphones can be obtained by Generalized Cross Correlation (GCC) method, and the least squares method was used to determine the Direction of Arrival (DoA) of the shock wave and muzzle blast. Finally, we simplify the existing Whitman model and a triangle model was established according to the acoustic characteristics of gunfire incident. By combining with the triangle model, simplified Whitman model, detected shock wave parameters, the time difference between the two kinds of sounds and the DoA,sniper position and bullet parameters can be effectively calculated. Finally, the experimental results validate the effectiveness and robustness of this algorithm.