The Design And Implement Of Missing And Begging Children Retrieval System Based On Hadoop Platform

This thesis comes from the Chinese social relief foundation projects which named"Crack Down on the Abduct and Traffic Children in Micro Blog". In recent years, with therising number of the lost, abducted and abandon children in China, kinds of people searchwebsites appear in the society. However, most of these websites just provide uploadfunction, after that they rely on the enthusiastic people confirm the lost people information.Therefore this paper hopes that through the face recognition technology, it can make gooduse of missing people images for retrieval. While implementing the actual face recognitionalgorithm, as face pictures too many, the information processing efficiency become slowly,directly affect influence the opportunity of rescuing the missing children. Under thissituation, using the Hadoop distributed platform can efficiently solve the huge dataprocessing problem and hope to make a contribution to rescuing the missing children.The most important works in this thesis are as follow. First, this paper designs andimplements a missing and begging children retrieval system which is based on Hadoopplatform. Second, use the Map/Reduce computing model to accelerate the speed in facerecognition. Last, this paper use cloud computing database to store the face feature and therelated person information. As a result, this paper not only solves the huge data storageproblem, but also let the face recognition system get the face feature more quickly.There are two main contributions in this thesis. One of them is setting up the missingand begging children retrieval system which is based on Hadoop platform. And then, thesystem can add other face recognition algorithm with a low price to adapt different imagefeature in different scenes. Another is using the face recognition system to find lostchildren by building a lost child database, responds of prevent the Abduct and traffic child. This paper designs and implements the missing and begging children retrieval systemwhich is based on Hadoop platform. The system is according to the software engineeringdevelopment process and specification. This paper designs the system by requirementsanalysis, general design, database design and detailed design, etc. In the face of featureextraction algorithm, this paper use JNI technology to accelerate the feature extractionefficiency and improve the system reusability. This system has being online and in thedevelopment of continuous improvement.