Research On Multi-biometrical Identity Recognition Method Based On SVM

In today’s society which information has highly developed, biometric technologyprovide an effective way for identification and subject to the attention of more andmore researchers. Compared to a single feature recognition, multi-biometrical identityrecognition has the following characteristics:1)greater reliability which can improvethe performace and fault tolerance of the system and reduce noise;2)broaderapplicability which make up the single-frature’s shortcomings that it may be notsuitable for some people;3)stronger security and so on. It has a very broad applicationprospects.The Support Vector Machine is a new tool for machine learning based onoptimization methods of staristical learning theory. It has the ability to solve highdimension, nonlinear and small samples. It has been widely used in the field ofbioinformatics, artificial intelligence and so on. Using the SVM for biometricidentification has a very important significance.The main work of this paper is:1. Because of the insufficient of the primitive face images, fingerprint imagesand iris images, these images have to be processed to eliminate noise and extractfeatures. Use SVM to build the classifier and to train and test the features. Theexperments has feasibility and a high recognition rate.2. An improved Gaussian function is researed to build the kernel function. Selectthe appropriate parameters to improve the learning and classification abilities of theGaussian kernel function. Its legitimacy has been proved. The improved Gaussiankernel is uesd for biometric experiment. Compared with the RBF kernel function, itimproved the performace of the system. 3. A decision level fusion method of the three single feature recognition isstudied. Pairwise fusion and fusion of the three are separate experiments. The threebiometric databases are simulately correspond to establish a multi-feature database.The experiments suggest the practicality, effectiveness and scalabitity of themulti-feature fusion authentication.