Abstract: For the location problem of unexploded ordnance (UXO) in the processes of conventional weapon test, military training and exercises, a low cost UXO detection method of the terminal ballistics is presented. According to the size of the impact area , many low cost acoustic sensors are installed in shooting range. The measuring range of these sensors covers the whole impact area. As for the signals of aerodynamic noise and the landing sound, which are captured by each acoustic sensor, the computation is implemented to used FFT and Laplace wavelet to denoise and enhance the original acoustic signal; use the short time energy, short time amplitude, and short time zero-crossing rate to detect the endpoint of these signals above; use the wavelet packet to collect the features of the enhanced signal; and use the minimized distance-based threshold criterion to identify the aerodynamic noise and the landing sound. The test results of UXO location in the shooting range show that the proposed technique can be used for UXO location and the location accuracy can reach 10 m.

Key words: ordnance science and technology, unexploded ordnance, endpoint detection, wavelet analysis, feature identification, acoustic location