Personal Identification Technology Based On Integrated Fingerprint And Handwritten Signature

Biological characteristics of identification technology is of very good accuracy, security and reliability, therefore, it develops rapidly in various fields and is gradually being applied to our daily life. However, it can't resolve its own difficulties encountered in the development, multi-biological characteristics of identification technology emerge as the times require.Based on the described above, this paper presents a multi-biological characteristics of identification method which integrates fingerprint with handwritten signature. The integration method improves the accuracy rate and security of identity recognition.First of all, the paper introduces the method of Gabor filtering into the algorithm of fingerprint identification. According to the characteristics of fingerprint ridge, determining fingerprint reference point using the filter processing. Then feature extraction region of fingerprint images is determined, and its normalization and Gabor filtering are also done. Finally fingerprint features are abstracted from the filtered fingerprint images and matched. The identification algorithm of fingerprint characteristics is tested by the fingerprint images in FVC2004 fingerprint database, the experiment results show that the algorithm is of good accuracy and reliability.Secondly, this paper further studies the identification algorithm of handwritten signature, using DTW signature feature matching method. To solve the shortage of single matching template, this paper presents a new matching template of handwritten signature and threshold determination method, and using DTW algorithm matches handwritten signature features. In the end, According to the contrastive analysis of experiment results, the identification algorithm has great improvement not only on recognition accuracy, but also on identification stability.On the basis of study above, this paper presents an identification technology of integrated fingerprint and handwritten signature, which mainly integrates from feature level and decision-making level. The integrated approach of feature level adopts Principal Component Analysis to reduce the dimension of feature, in order to resolve the difficulties raised by high feature dimension, the integrated approach of decision-making level taking advantage of Bayes decision theory establishes the mathematical model of decision-making optimization, and which is optimized by DL-BPSO. According to the comparative analysis of experiment results, the integrated approach can effectively improve identity recognition accuracy and security performance.