Wavelet Analysis Applied To The Research On The Detection Of Drug/Explosive

Energy-dispersive X-ray analysis technology is a significant technique to identify substance. It has been penetrated into multi-disciplines, such as branches of Detection Technology, Electronics Technology, Computers and so on. Since it has the characteristics of low cost, multi-elements measuring quickly etc, now it has been widely applied to industry, agriculture and bio-medical. In this paper, we substitute synchrotron radiation light source with an ordinary X-ray machine to set up a novel energy-dispersive X-ray device, which irradiates drug and explosive by low radiation dosage in the range of little-angle. Thus we could seek for the characteristical scattering peaks of X-ray energy spectrum signal obtained by energy-dispersive X-ray device to identify material composition.The study originates from National Nature Science Foundation, namely the principle and application of high-speed X-ray cone-beam imaging and the novel scattering method to detect drug and explosive (Foundation No is 10635070). However, owing to low radiation dosage and low energy resolving power of the detector, energy spectrum signal is polluted by noise so that the baseline of energy spectrum signal has some obvious drift and some scattering peaks widen and interference peaks mixed. Therefore, we couldn't effectively recognize sample composition. Consequently, in the paper, we utilize wavelet technology with many merits to process energy spectrum signal. On the base of such different de-noising algorithm, a novel adaptive wavelet spatial correlative de-noising algorithm is put forward, which could effectively deal with the noise spectrum. According to the multi-resolution characteristic of wavelet, it would divide energy spectrum signal into different frequency components. The paper uses an algorithm of repeatedly cutting peaks to obtain the approximation of the background so that the background continuum could be effectively substracted. Owing to scattering peaks of sample and other impurities overlapped, a new method called second-order spline convolution algorithm is applied to separate overlapped X-ray peaks. The corresponding experimental results proof good. Finally, making use of wavelet singularity analysis, we could effectively get all characteristical scattering peaks of energy spectrum signal by peeking for the key points of wavelet transform peaks.Wavelet analysis technology is introduced into the processing of energy spectrum signal of drug and explosive. It improves the analysis accuracy of material composition, and provides theoretical foundation and practical guidance for further study of wavelet analysis applied to signal processing in the energy spectrum.