Font Size: a A A

X-ray Scattering Detection And Identification Algorithm Of Precursor Chemicals

Posted on:2012-09-29Degree:MasterType:Thesis
Country:ChinaCandidate:Y ZhangFull Text:PDF
GTID:2178330338992162Subject:Detection Technology and Automation
Abstract/Summary:PDF Full Text Request
Recently, the smuggling and trafficking of precursor chemicals and drugs is becoming increasingly serious. The current gas chromatography and Raman spectroscopy detection technologies cannot simultaneously have the all the properties for rapid, accurate and non-destructive testing. X-ray coherent scattering technique is an important material recognition technology with the advantages of non-destructive, non-invasive, low-cost, high-resolution and so on. It is widely used in industry,agriculture,bio-medicine and many other fields. With no need to destroy the packaging, the strong ability to penetrate through packaging of X-ray is used to analyze the intrinsic structure of the packaged sample within the level of the molecular structure.This research originates from National Natural Science Fund: Principles and Applications of high speed three-dimensional image of X-ray scattering and new methods for detecting drugs and explosives (Code: 10635070). In this paper, the X-ray coherent scattering test platform which is independently developed by our group is used to detect a variety of precursor chemicals. Due to the shape similarity of scattering spectra of several samples, they cannot be directly classified. Therefore, the method based on singular value decomposition (SVD) is introduced to analyze the scattering spectra. Based on the research of existing feature extraction algorithm, a new feature extraction algorithm which will directly use the primary scattering spectra is developed in this paper. The extracted characteristics by the algorithm can be used to classify and differentiate some kinds of precursor chemicals (including the samples whose scattering spectra are very similar).After the research of classification of precursor chemicals, detecting and identification of drugs/explosives packaged in human model is studied using recognition algorithm based on support vector machine (SVM). A good identification result is achieved. This is a meaningful exploratory research for using X-ray coherent scattering technology in the identification of drugs packaged in actual human bodies.
Keywords/Search Tags:X-ray coherent scattering, precursor chemical, singular value decomposition, drugs/explosives, support vector machine
PDF Full Text Request
Related items