Feature Level Fusion Based On Kernel Method--And Its Application On Handmetric Recognition

Biometrics is the most secure and convenient way to satisfy the requirements for identity digitalization and virtualization in the coming network society, which refers to the automatic identification of an individual by using certain physiological or behavioral traits associated with the person.Handmetric Recognition technologies become the hot focus in biometrics identification currently for its advantages。And According to the levels established by international literature, information fusion can occur at the data,feature,score or decision level.And the paper is just completed under the project"Research of Feature level information fusion theory and its application",funded by the National Natural Science Fund Committee.With the idea of feature level information fusion,we make deeply research on Kernel method,and Approaches proposed to improve the performance of it.The main works of the paper are as fellows:1,with the begin,we introduce the Obtainment and pre-processing of hand image. Then an approache on Palmprint interception base on Binary mosaic.And we also make improving on key points location,that solves the problem of inaccuracy of region-of-interest(ROI)extract.2,Based on the investigation of the existing feature level fusion algorithms and the idea of well studied Kernel method,new Kernel method called KL kernel and L^2 kernel are proposed.To solving the problem of the limitations of a single kernel function,Hyperkernel function is proposed.Then we give out experiments on palmprint, to prove the validity of new kernel function.3,A more reliable algorithm based on kernel PCA(KPCA) is deducted form RMF for handmetric recognition,to resolve the problem of the inconsistency of dimensions, numerical difference and so on.With the fusion feature of the linear Principal Component Analysis(PCA) and the Principal Component Analysis based kernel,its application on the palmprint identification can get higher identification rate.The experiments base on palmprint and four different knuckle prints using the optimized KPCA based fusion method.The evaluation results show that the performance of proposed algorithm is the best in altogether methods.