Font Size: a A A

The Design And Realization Of Monitor System Based On Face Recognition

Posted on:2010-09-05Degree:MasterType:Thesis
Country:ChinaCandidate:G L ZhangFull Text:PDF
GTID:2178360278959532Subject:Signal and Information Processing
Abstract/Summary:PDF Full Text Request
The rapid improvements of computer and recognition technologies have brought the biometric recognition technology,like recognition of palm, fingerprint,voice,iris,face,into actual and accurate use.Among these novel technologies,face recognition is superior to others.In view of the limitation of independent operation of monitoring system and theft proof system,this system is set up in C++,by integrating development software and combining with the advanced biologic identification technology,and communication technology.By fully utilizing the object-oriented design technology,we develop an all-around function face identification system.The structure and program realization scheme of the system and the composition and character of the system are introduced in this paper.The main contents and originalities in this thesis are listed as follows:1,Describes significance and the research background of the subject.At the same time,does an analysis of the general summary both at home and abroad to;gives the needs analysis,functional analysis,functional structure of a face recognition-based security monitoring system and the work flow of the system;2,Detailed analysis of the basic theory on CAMSHIFT,proposes that the CAMSHIFT based on adaptive Kalman filtering and face detection tracking algorithm.By comparing the traditional algorithm CAMSHIF,the experimental results show that the algorithm proposed in this paper has a higher detection accuracy and faster speed of real-time tracking;3,Study the characteristics of the wavelet transform,the current popular AdaBoost algorithm is applied to the wavelet feature selection table for the extraction of the identification of the most powerful features,and finally through the discriminant analysis of the characteristics of implementation to identify and put forward an approach based on AdaBoost with the characteristics of wavelet transform method of linear discriminant analysis;4,List the implementation details of security monitoring system based on face recognition,such as database design,flow register,the design of control circuit.Finally,based on the algorithm proposed by the previous section and solutions to let people to face a safety monitoring system access control, attendance,monitor the application of three typical prototype system design and systems interface.
Keywords/Search Tags:face recognition, adaboost algorithm, CAMSHIFT algorithm, wavelets
PDF Full Text Request
Related items