| X-ray has been widely used in the field of substance recognition because of its powerful penetrating ability and unique physical reaction mechanism.X-ray transmission technology is mainly used for the detection of objects by imaging the structure of the detected object.X-ray absorption spectroscopy mainly uses synchrotron radiation as a light source,and it is generally used for chemical element analysis.This article intends to use X-ray tube as the light source and use the photon counting detector to obtain the X-ray absorption spectrum of substance,and combine X-ray transmission technology to recognize substance.This paper compares the advantages and disadvantages of various types of substance recognition methods,and studies X-ray absorption spectral analytical theory in detail.In this paper,experiments were performed on various plastic samples and various porcine tissue samples based on X-ray spectrum detection system.The X-ray absorption spectra of each sample were analyzed and it was found that different substances have certain spectral differences.But the spectral differences of some samples are small,and it is difficult to distinguish these spectra from visual observation.The principal component analysis method and artificial neural network algorithm were introduced and derived in terms of substance recognition methods.The principal component analysis method was used to extract the spectral features of the X-ray absorption spectra of samples,and it was found that the spectral features of same type of samples could be clustered together better.Then the neural network model was established to recognize the spectral features.The recognition rates of 15 kinds of plastic samples and 6 kinds of porcine tissue samples reached 99.33%and 90.22%,respectively.Experiments have proved that the use of X-ray absorption spectroscopy can recognize substances,which has good application prospects in medical diagnosis,safety detection,non-destructive testing and other fields. |
PDF Full Text Request