Peak detection of fragments velocity measurement in laser screen

Abstract

In laser screen fragments velocity measurement system, the accurate time of the fragment across the laser screen affects the velocity precision. The traditional peak detection using numerical differentiation of signal directly can cause the wrong detection of the noises, and the direct peak-detection method can cause signal error detection because of effects of noises also. A method of the peak detection based on wavelet transformed modulus maximum is proposed. Firstly appropriate wavelet basis is selected and used to wavelet transform of fragment signal. Secondly peak detection and denoising is realized based on the relationship of wavelet modulus maxima, Lipschitz exponent and singularity of fragment signal. Peaks can be detected by the method of the wavelet transform modulus maxima method correctly. The 96.9% recognition rate of fragments can be obtained by using the method through a large number of test data processing.