Design Of Explosive Residue Analysis System

The bombing is a very serious violent crime,which does great harm to public security and social stability.At present,the physical evidence identification of explosion cases in China needs experts to carry out comprehensive identification on the basis of various inspection data.At the same time,there is also a lack of analysis software platform that is easy to operate and easy to be deployed in various places.This paper needs to design an analysis software system that meets the actual needs of the public security department.At present,with the popularization of broadband network and the development of modern browser technology,more and more software applications have given up the form of desktop client and turned to browser environment.Therefore,this thesis designs an analysis software system of explosive residue based on Web technology according to the requirements of relevant units involved in the research project.The results include:In terms of designing the system,according to the task requirements,this thesis describes the overall design of the comprehensive analysis system for explosive residues.The browser/server architecture was adopted to facilitate the promotion,deployment and use of the system by relevant units,so as to give full play to its practical value.In the aspect of front-end design,this paper designs the front-end module and program of the analysis system.The front-end program was developed by using Vue and other Web technologies.The front-end program was run in the browser,so that the system platform could be flexibly accessed in various places.In terms of back-end and database design,based on Django framework and MySQL database,this paper design the back-end module and program of the analysis system,as well as the database part containing components and shapes of sample,which can correctly and effectively respond to the requests issued by the front-end and complete various system functions.In terms of core recognition algorithm embedded in system,GrabCut-based segmentation algorithm and a recognition algorithm based on feature points/components structure relation are integrated in the system.The experimantal results demonstrate the effectiveness of our methods.This system has provided a service after tress tests.